Upload folder using huggingface_hub

3b-w-cot+/README.md

@@ -0,0 +1,163 @@
+---
+library_name: transformers
+tags:
+- generated_from_trainer
+datasets:
+- train-mb.jsonl
+model-index:
+- name: outputs/out
+  results: []
+---
+
+<!-- This model card has been generated automatically according to the information the Trainer had access to. You
+should probably proofread and complete it, then remove this comment. -->
+
+[<img src="https://raw.githubusercontent.com/axolotl-ai-cloud/axolotl/main/image/axolotl-badge-web.png" alt="Built with Axolotl" width="200" height="32"/>](https://github.com/axolotl-ai-cloud/axolotl)
+<details><summary>See axolotl config</summary>
+
+axolotl version: `0.8.0.dev0`
+```yaml
+base_model: ckpt/3b-w-cot/checkpoint-747
+model_type: AutoModelForCausalLM
+tokenizer_type: AutoTokenizer
+trust_remote_code: false
+
+load_in_8bit: false
+load_in_4bit: false
+strict: false
+
+output_dir: ./outputs/out
+chat_template: qwen_25
+datasets:
+  - path: train-mb.jsonl
+    type: chat_template
+    field_messages: messages
+    message_field_role: role
+    message_field_content: content
+    roles:
+      system:
+        - system
+      user:
+        - user
+      assistant:
+        - assistant
+
+dataset_prepared_path: last_run_prepared
+val_set_size: 0.005
+output_dir: ./outputs/out
+eval_sample_packing: False
+
+sequence_len: 8192
+sample_packing: False
+pad_to_sequence_len: False
+
+wandb_project: mergedbench
+wandb_entity:
+wandb_watch:
+wandb_name:
+wandb_log_model:
+# hub_model_id: amphora/merged-bench-qwen-full
+
+plugins:
+  - axolotl.integrations.liger.LigerPlugin
+liger_rope: true
+liger_rms_norm: true
+liger_swiglu: true
+liger_fused_linear_cross_entropy: true
+
+gradient_accumulation_steps: 4
+micro_batch_size: 8
+eval_batch_size: 4
+num_epochs: 3
+optimizer: paged_adamw_8bit
+lr_scheduler: cosine
+learning_rate: 2e-5
+
+train_on_inputs: false
+group_by_length: false
+bf16: auto
+fp16:
+tf32: false
+
+gradient_checkpointing: true
+gradient_checkpointing_kwargs:
+  use_reentrant: false
+early_stopping_patience:
+resume_from_checkpoint:
+logging_steps: 1
+xformers_attention:
+flash_attention: true
+
+warmup_steps: 30
+evals_per_epoch: 3
+eval_max_new_tokens: 128
+eval_table_size:
+saves_per_epoch: 1
+debug:
+deepspeed: deepspeed_configs/zero1.json
+weight_decay: 0.01
+fsdp:
+fsdp_config:
+special_tokens:
+```
+
+</details><br>
+
+# outputs/out
+
+This model was trained from scratch on the train-mb.jsonl dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.2554
+
+## Model description
+
+More information needed
+
+## Intended uses & limitations
+
+More information needed
+
+## Training and evaluation data
+
+More information needed
+
+## Training procedure
+
+### Training hyperparameters
+
+The following hyperparameters were used during training:
+- learning_rate: 2e-05
+- train_batch_size: 8
+- eval_batch_size: 4
+- seed: 42
+- distributed_type: multi-GPU
+- num_devices: 2
+- gradient_accumulation_steps: 4
+- total_train_batch_size: 64
+- total_eval_batch_size: 8
+- optimizer: Use OptimizerNames.PAGED_ADAMW_8BIT with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
+- lr_scheduler_type: cosine
+- lr_scheduler_warmup_steps: 30
+- num_epochs: 3.0
+
+### Training results
+
+| Training Loss | Epoch  | Step | Validation Loss |
+|:-------------:|:------:|:----:|:---------------:|
+| 0.477         | 0.0041 | 1    | 0.8367          |
+| 0.2339        | 0.3357 | 82   | 0.2754          |
+| 0.2318        | 0.6714 | 164  | 0.2667          |
+| 0.1688        | 1.0041 | 246  | 0.2618          |
+| 0.1467        | 1.3398 | 328  | 0.2673          |
+| 0.1605        | 1.6755 | 410  | 0.2577          |
+| 0.1206        | 2.0082 | 492  | 0.2539          |
+| 0.1022        | 2.3439 | 574  | 0.2587          |
+| 0.0881        | 2.6796 | 656  | 0.2554          |
+
+
+### Framework versions
+
+- Transformers 4.49.0
+- Pytorch 2.5.1+cu121
+- Datasets 3.2.0
+- Tokenizers 0.21.0

3b-w-cot+/added_tokens.json

@@ -0,0 +1,24 @@
+{
+  "</tool_call>": 151658,
+  "<tool_call>": 151657,
+  "<|box_end|>": 151649,
+  "<|box_start|>": 151648,
+  "<|endoftext|>": 151643,
+  "<|file_sep|>": 151664,
+  "<|fim_middle|>": 151660,
+  "<|fim_pad|>": 151662,
+  "<|fim_prefix|>": 151659,
+  "<|fim_suffix|>": 151661,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644,
+  "<|image_pad|>": 151655,
+  "<|object_ref_end|>": 151647,
+  "<|object_ref_start|>": 151646,
+  "<|quad_end|>": 151651,
+  "<|quad_start|>": 151650,
+  "<|repo_name|>": 151663,
+  "<|video_pad|>": 151656,
+  "<|vision_end|>": 151653,
+  "<|vision_pad|>": 151654,
+  "<|vision_start|>": 151652
+}

3b-w-cot+/checkpoint-244/added_tokens.json

@@ -0,0 +1,24 @@
+{
+  "</tool_call>": 151658,
+  "<tool_call>": 151657,
+  "<|box_end|>": 151649,
+  "<|box_start|>": 151648,
+  "<|endoftext|>": 151643,
+  "<|file_sep|>": 151664,
+  "<|fim_middle|>": 151660,
+  "<|fim_pad|>": 151662,
+  "<|fim_prefix|>": 151659,
+  "<|fim_suffix|>": 151661,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644,
+  "<|image_pad|>": 151655,
+  "<|object_ref_end|>": 151647,
+  "<|object_ref_start|>": 151646,
+  "<|quad_end|>": 151651,
+  "<|quad_start|>": 151650,
+  "<|repo_name|>": 151663,
+  "<|video_pad|>": 151656,
+  "<|vision_end|>": 151653,
+  "<|vision_pad|>": 151654,
+  "<|vision_start|>": 151652
+}

3b-w-cot+/checkpoint-244/config.json

@@ -0,0 +1,28 @@
+{
+  "_name_or_path": "ckpt/3b-w-cot/checkpoint-747",
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "eos_token_id": 151645,
+  "hidden_act": "silu",
+  "hidden_size": 2048,
+  "initializer_range": 0.02,
+  "intermediate_size": 11008,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 70,
+  "model_type": "qwen2",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 36,
+  "num_key_value_heads": 2,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.49.0",
+  "use_cache": false,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}

3b-w-cot+/checkpoint-244/generation_config.json

@@ -0,0 +1,14 @@
+{
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "repetition_penalty": 1.05,
+  "temperature": 0.7,
+  "top_k": 20,
+  "top_p": 0.8,
+  "transformers_version": "4.49.0"
+}

3b-w-cot+/checkpoint-244/global_step244/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ea41c8f2726b36ff0db03ba02c1c7f67d206be3993e17812e702273aae19dd02
+size 9306058322

3b-w-cot+/checkpoint-244/global_step244/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fbec0a25e4ee80ff214f60de99f85cd1b59bd87a46e5d9dd13be56139577b3e3
+size 9306060690

3b-w-cot+/checkpoint-244/global_step244/mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ee7c23d50026e8003e0359f5a0ed45475e0bafa887d37eeb0dbb473a365e6c24
+size 6171993592

3b-w-cot+/checkpoint-244/latest

@@ -0,0 +1 @@
+global_step244

3b-w-cot+/checkpoint-244/model-00001-of-00002.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:902bf38b67efd1e0f84e8b3cd2e6f0ce844ae92f29b646a5cf9bb1560fde4d55
+size 4957560304

3b-w-cot+/checkpoint-244/model-00002-of-00002.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:04e7e4bcb562e8498e019b7d3ed72400d6a30fd934801232970e3bed8f77a698
+size 1836696752

3b-w-cot+/checkpoint-244/model.safetensors.index.json

@@ -0,0 +1,442 @@
+{
+  "metadata": {
+    "total_size": 6794207232
+  },
+  "weight_map": {
+    "lm_head.weight": "model-00002-of-00002.safetensors",
+    "model.embed_tokens.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.28.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.28.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.28.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.28.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.28.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.28.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.28.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.29.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.30.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.norm.weight": "model-00002-of-00002.safetensors"
+  }
+}

3b-w-cot+/checkpoint-244/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a9affc1541e7e94c18354d5173bc55400c5f07faf3d080c6d453d48e7a8d6ac3
+size 14512

3b-w-cot+/checkpoint-244/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4748c3ebf0e4c051c58b92e4a8c5b87cdb39d55cfdc2aec81a1baef0f02fc113
+size 14512

3b-w-cot+/checkpoint-244/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5cb186d02e42c19d881269361281b0d1dc724284e39baf6809ced6fd93070319
+size 1064

3b-w-cot+/checkpoint-244/special_tokens_map.json

@@ -0,0 +1,31 @@
+{
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "eos_token": {
+    "content": "<|im_end|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

3b-w-cot+/checkpoint-244/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9c5ae00e602b8860cbd784ba82a8aa14e8feecec692e7076590d014d7b7fdafa
+size 11421896

3b-w-cot+/checkpoint-244/tokenizer_config.json

@@ -0,0 +1,208 @@
+{
+  "add_bos_token": false,
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151645": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151646": {
+      "content": "<|object_ref_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151647": {
+      "content": "<|object_ref_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151648": {
+      "content": "<|box_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151649": {
+      "content": "<|box_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151650": {
+      "content": "<|quad_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151651": {
+      "content": "<|quad_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151652": {
+      "content": "<|vision_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151653": {
+      "content": "<|vision_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151654": {
+      "content": "<|vision_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151655": {
+      "content": "<|image_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151656": {
+      "content": "<|video_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151657": {
+      "content": "<tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151658": {
+      "content": "</tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151659": {
+      "content": "<|fim_prefix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151660": {
+      "content": "<|fim_middle|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151661": {
+      "content": "<|fim_suffix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151662": {
+      "content": "<|fim_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151663": {
+      "content": "<|repo_name|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151664": {
+      "content": "<|file_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    }
+  },
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "bos_token": null,
+  "chat_template": "{%- if tools %}\n    {{- '<|im_start|>system\\n' }}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- messages[0]['content'] }}\n    {%- else %}\n        {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}\n    {%- endif %}\n    {{- \"\\n\\n# Tools\\n\\nYou may call one or more functions to assist with the user query.\\n\\nYou are provided with function signatures within <tools></tools> XML tags:\\n<tools>\" }}\n    {%- for tool in tools %}\n        {{- \"\\n\" }}\n        {{- tool | tojson }}\n    {%- endfor %}\n    {{- \"\\n</tools>\\n\\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\\n<tool_call>\\n{\\\"name\\\": <function-name>, \\\"arguments\\\": <args-json-object>}\\n</tool_call><|im_end|>\\n\" }}\n{%- else %}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- '<|im_start|>system\\n' + messages[0]['content'] + '<|im_end|>\\n' }}\n    {%- else %}\n        {{- '<|im_start|>system\\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\\n' }}\n    {%- endif %}\n{%- endif %}\n{%- for message in messages %}\n    {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) or (message.role == \"assistant\" and not message.tool_calls) %}\n        {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {{- '<|im_start|>' + message.role }}\n        {%- if message.content %}\n            {{- '\\n' + message.content }}\n        {%- endif %}\n        {%- for tool_call in message.tool_calls %}\n            {%- if tool_call.function is defined %}\n                {%- set tool_call = tool_call.function %}\n            {%- endif %}\n            {{- '\\n<tool_call>\\n{\"name\": \"' }}\n            {{- tool_call.name }}\n            {{- '\", \"arguments\": ' }}\n            {{- tool_call.arguments | tojson }}\n            {{- '}\\n</tool_call>' }}\n        {%- endfor %}\n        {{- '<|im_end|>\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != \"tool\") %}\n            {{- '<|im_start|>user' }}\n        {%- endif %}\n        {{- '\\n<tool_response>\\n' }}\n        {{- message.content }}\n        {{- '\\n</tool_response>' }}\n        {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n            {{- '<|im_end|>\\n' }}\n        {%- endif %}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {{- '<|im_start|>assistant\\n' }}\n{%- endif %}\n",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": {},
+  "model_max_length": 131072,
+  "pad_token": "<|endoftext|>",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

3b-w-cot+/checkpoint-244/trainer_state.json

@@ -0,0 +1,1765 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9989764585465711,
+  "eval_steps": 82,
+  "global_step": 244,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0040941658137154556,
+      "grad_norm": 8.007163047790527,
+      "learning_rate": 6.666666666666667e-07,
+      "loss": 0.477,
+      "step": 1
+    },
+    {
+      "epoch": 0.0040941658137154556,
+      "eval_loss": 0.8367487192153931,
+      "eval_runtime": 4.4844,
+      "eval_samples_per_second": 17.617,
+      "eval_steps_per_second": 2.23,
+      "step": 1
+    },
+    {
+      "epoch": 0.008188331627430911,
+      "grad_norm": 8.800883293151855,
+      "learning_rate": 1.3333333333333334e-06,
+      "loss": 0.6282,
+      "step": 2
+    },
+    {
+      "epoch": 0.012282497441146366,
+      "grad_norm": 9.306445121765137,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.6202,
+      "step": 3
+    },
+    {
+      "epoch": 0.016376663254861822,
+      "grad_norm": 10.017245292663574,
+      "learning_rate": 2.666666666666667e-06,
+      "loss": 0.6161,
+      "step": 4
+    },
+    {
+      "epoch": 0.02047082906857728,
+      "grad_norm": 9.14148235321045,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.6081,
+      "step": 5
+    },
+    {
+      "epoch": 0.02456499488229273,
+      "grad_norm": 8.813340187072754,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.5697,
+      "step": 6
+    },
+    {
+      "epoch": 0.028659160696008188,
+      "grad_norm": 6.7533745765686035,
+      "learning_rate": 4.666666666666667e-06,
+      "loss": 0.4834,
+      "step": 7
+    },
+    {
+      "epoch": 0.032753326509723645,
+      "grad_norm": 6.1987481117248535,
+      "learning_rate": 5.333333333333334e-06,
+      "loss": 0.4857,
+      "step": 8
+    },
+    {
+      "epoch": 0.0368474923234391,
+      "grad_norm": 2.4827005863189697,
+      "learning_rate": 6e-06,
+      "loss": 0.3713,
+      "step": 9
+    },
+    {
+      "epoch": 0.04094165813715456,
+      "grad_norm": 2.163064956665039,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.3284,
+      "step": 10
+    },
+    {
+      "epoch": 0.04503582395087001,
+      "grad_norm": 1.9997942447662354,
+      "learning_rate": 7.333333333333333e-06,
+      "loss": 0.3289,
+      "step": 11
+    },
+    {
+      "epoch": 0.04912998976458546,
+      "grad_norm": 2.7956204414367676,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.3208,
+      "step": 12
+    },
+    {
+      "epoch": 0.05322415557830092,
+      "grad_norm": 2.8886733055114746,
+      "learning_rate": 8.666666666666668e-06,
+      "loss": 0.3123,
+      "step": 13
+    },
+    {
+      "epoch": 0.057318321392016376,
+      "grad_norm": 2.217071771621704,
+      "learning_rate": 9.333333333333334e-06,
+      "loss": 0.2881,
+      "step": 14
+    },
+    {
+      "epoch": 0.06141248720573183,
+      "grad_norm": 1.9985229969024658,
+      "learning_rate": 1e-05,
+      "loss": 0.283,
+      "step": 15
+    },
+    {
+      "epoch": 0.06550665301944729,
+      "grad_norm": 1.8881174325942993,
+      "learning_rate": 1.0666666666666667e-05,
+      "loss": 0.2616,
+      "step": 16
+    },
+    {
+      "epoch": 0.06960081883316274,
+      "grad_norm": 1.9551236629486084,
+      "learning_rate": 1.1333333333333334e-05,
+      "loss": 0.2894,
+      "step": 17
+    },
+    {
+      "epoch": 0.0736949846468782,
+      "grad_norm": 1.8677968978881836,
+      "learning_rate": 1.2e-05,
+      "loss": 0.2328,
+      "step": 18
+    },
+    {
+      "epoch": 0.07778915046059365,
+      "grad_norm": 1.9170935153961182,
+      "learning_rate": 1.2666666666666667e-05,
+      "loss": 0.2577,
+      "step": 19
+    },
+    {
+      "epoch": 0.08188331627430911,
+      "grad_norm": 2.189279794692993,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.2555,
+      "step": 20
+    },
+    {
+      "epoch": 0.08597748208802457,
+      "grad_norm": 3.202075242996216,
+      "learning_rate": 1.4e-05,
+      "loss": 0.2647,
+      "step": 21
+    },
+    {
+      "epoch": 0.09007164790174002,
+      "grad_norm": 2.774186372756958,
+      "learning_rate": 1.4666666666666666e-05,
+      "loss": 0.2531,
+      "step": 22
+    },
+    {
+      "epoch": 0.09416581371545547,
+      "grad_norm": 2.0601961612701416,
+      "learning_rate": 1.5333333333333334e-05,
+      "loss": 0.2654,
+      "step": 23
+    },
+    {
+      "epoch": 0.09825997952917093,
+      "grad_norm": 1.781900405883789,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.2358,
+      "step": 24
+    },
+    {
+      "epoch": 0.1023541453428864,
+      "grad_norm": 1.8549216985702515,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.2362,
+      "step": 25
+    },
+    {
+      "epoch": 0.10644831115660185,
+      "grad_norm": 2.1376802921295166,
+      "learning_rate": 1.7333333333333336e-05,
+      "loss": 0.2334,
+      "step": 26
+    },
+    {
+      "epoch": 0.1105424769703173,
+      "grad_norm": 2.134582281112671,
+      "learning_rate": 1.8e-05,
+      "loss": 0.267,
+      "step": 27
+    },
+    {
+      "epoch": 0.11463664278403275,
+      "grad_norm": 1.6425909996032715,
+      "learning_rate": 1.866666666666667e-05,
+      "loss": 0.2184,
+      "step": 28
+    },
+    {
+      "epoch": 0.1187308085977482,
+      "grad_norm": 1.601938009262085,
+      "learning_rate": 1.9333333333333333e-05,
+      "loss": 0.256,
+      "step": 29
+    },
+    {
+      "epoch": 0.12282497441146366,
+      "grad_norm": 1.626160740852356,
+      "learning_rate": 2e-05,
+      "loss": 0.2548,
+      "step": 30
+    },
+    {
+      "epoch": 0.1269191402251791,
+      "grad_norm": 1.6894042491912842,
+      "learning_rate": 1.999989986294826e-05,
+      "loss": 0.2438,
+      "step": 31
+    },
+    {
+      "epoch": 0.13101330603889458,
+      "grad_norm": 1.6300302743911743,
+      "learning_rate": 1.9999599453798523e-05,
+      "loss": 0.2273,
+      "step": 32
+    },
+    {
+      "epoch": 0.13510747185261002,
+      "grad_norm": 1.6301401853561401,
+      "learning_rate": 1.999909877856721e-05,
+      "loss": 0.2539,
+      "step": 33
+    },
+    {
+      "epoch": 0.13920163766632548,
+      "grad_norm": 1.6974273920059204,
+      "learning_rate": 1.9998397847281548e-05,
+      "loss": 0.256,
+      "step": 34
+    },
+    {
+      "epoch": 0.14329580348004095,
+      "grad_norm": 1.5356749296188354,
+      "learning_rate": 1.9997496673979375e-05,
+      "loss": 0.2278,
+      "step": 35
+    },
+    {
+      "epoch": 0.1473899692937564,
+      "grad_norm": 1.6304699182510376,
+      "learning_rate": 1.9996395276708856e-05,
+      "loss": 0.2488,
+      "step": 36
+    },
+    {
+      "epoch": 0.15148413510747186,
+      "grad_norm": 1.6241912841796875,
+      "learning_rate": 1.999509367752813e-05,
+      "loss": 0.2407,
+      "step": 37
+    },
+    {
+      "epoch": 0.1555783009211873,
+      "grad_norm": 1.6954501867294312,
+      "learning_rate": 1.9993591902504854e-05,
+      "loss": 0.2279,
+      "step": 38
+    },
+    {
+      "epoch": 0.15967246673490276,
+      "grad_norm": 1.5775200128555298,
+      "learning_rate": 1.9991889981715696e-05,
+      "loss": 0.2443,
+      "step": 39
+    },
+    {
+      "epoch": 0.16376663254861823,
+      "grad_norm": 1.6417704820632935,
+      "learning_rate": 1.9989987949245725e-05,
+      "loss": 0.2498,
+      "step": 40
+    },
+    {
+      "epoch": 0.16786079836233367,
+      "grad_norm": 1.6866360902786255,
+      "learning_rate": 1.9987885843187717e-05,
+      "loss": 0.2496,
+      "step": 41
+    },
+    {
+      "epoch": 0.17195496417604914,
+      "grad_norm": 1.6979321241378784,
+      "learning_rate": 1.9985583705641418e-05,
+      "loss": 0.2721,
+      "step": 42
+    },
+    {
+      "epoch": 0.17604912998976457,
+      "grad_norm": 1.676047682762146,
+      "learning_rate": 1.9983081582712684e-05,
+      "loss": 0.2506,
+      "step": 43
+    },
+    {
+      "epoch": 0.18014329580348004,
+      "grad_norm": 1.7262601852416992,
+      "learning_rate": 1.998037952451255e-05,
+      "loss": 0.2371,
+      "step": 44
+    },
+    {
+      "epoch": 0.1842374616171955,
+      "grad_norm": 1.4194152355194092,
+      "learning_rate": 1.9977477585156252e-05,
+      "loss": 0.2619,
+      "step": 45
+    },
+    {
+      "epoch": 0.18833162743091095,
+      "grad_norm": 1.5654889345169067,
+      "learning_rate": 1.9974375822762117e-05,
+      "loss": 0.2299,
+      "step": 46
+    },
+    {
+      "epoch": 0.19242579324462641,
+      "grad_norm": 1.844489336013794,
+      "learning_rate": 1.9971074299450414e-05,
+      "loss": 0.2692,
+      "step": 47
+    },
+    {
+      "epoch": 0.19651995905834185,
+      "grad_norm": 1.5128370523452759,
+      "learning_rate": 1.9967573081342103e-05,
+      "loss": 0.2589,
+      "step": 48
+    },
+    {
+      "epoch": 0.20061412487205732,
+      "grad_norm": 1.5005507469177246,
+      "learning_rate": 1.9963872238557516e-05,
+      "loss": 0.2578,
+      "step": 49
+    },
+    {
+      "epoch": 0.2047082906857728,
+      "grad_norm": 1.5974067449569702,
+      "learning_rate": 1.9959971845214953e-05,
+      "loss": 0.2494,
+      "step": 50
+    },
+    {
+      "epoch": 0.20880245649948823,
+      "grad_norm": 1.5728641748428345,
+      "learning_rate": 1.9955871979429188e-05,
+      "loss": 0.2496,
+      "step": 51
+    },
+    {
+      "epoch": 0.2128966223132037,
+      "grad_norm": 1.5953929424285889,
+      "learning_rate": 1.9951572723309918e-05,
+      "loss": 0.2429,
+      "step": 52
+    },
+    {
+      "epoch": 0.21699078812691913,
+      "grad_norm": 1.7769081592559814,
+      "learning_rate": 1.9947074162960113e-05,
+      "loss": 0.252,
+      "step": 53
+    },
+    {
+      "epoch": 0.2210849539406346,
+      "grad_norm": 1.6964116096496582,
+      "learning_rate": 1.9942376388474282e-05,
+      "loss": 0.2651,
+      "step": 54
+    },
+    {
+      "epoch": 0.22517911975435004,
+      "grad_norm": 1.5599926710128784,
+      "learning_rate": 1.993747949393668e-05,
+      "loss": 0.2193,
+      "step": 55
+    },
+    {
+      "epoch": 0.2292732855680655,
+      "grad_norm": 1.524835467338562,
+      "learning_rate": 1.9932383577419432e-05,
+      "loss": 0.2361,
+      "step": 56
+    },
+    {
+      "epoch": 0.23336745138178097,
+      "grad_norm": 1.6240477561950684,
+      "learning_rate": 1.992708874098054e-05,
+      "loss": 0.2563,
+      "step": 57
+    },
+    {
+      "epoch": 0.2374616171954964,
+      "grad_norm": 1.6357301473617554,
+      "learning_rate": 1.9921595090661872e-05,
+      "loss": 0.2456,
+      "step": 58
+    },
+    {
+      "epoch": 0.24155578300921188,
+      "grad_norm": 1.5650979280471802,
+      "learning_rate": 1.991590273648702e-05,
+      "loss": 0.2512,
+      "step": 59
+    },
+    {
+      "epoch": 0.24564994882292732,
+      "grad_norm": 1.4107614755630493,
+      "learning_rate": 1.9910011792459086e-05,
+      "loss": 0.2539,
+      "step": 60
+    },
+    {
+      "epoch": 0.24974411463664278,
+      "grad_norm": 1.3979556560516357,
+      "learning_rate": 1.9903922376558432e-05,
+      "loss": 0.2348,
+      "step": 61
+    },
+    {
+      "epoch": 0.2538382804503582,
+      "grad_norm": 1.4066411256790161,
+      "learning_rate": 1.989763461074029e-05,
+      "loss": 0.2419,
+      "step": 62
+    },
+    {
+      "epoch": 0.2579324462640737,
+      "grad_norm": 1.533858060836792,
+      "learning_rate": 1.989114862093232e-05,
+      "loss": 0.252,
+      "step": 63
+    },
+    {
+      "epoch": 0.26202661207778916,
+      "grad_norm": 1.4007140398025513,
+      "learning_rate": 1.9884464537032103e-05,
+      "loss": 0.2379,
+      "step": 64
+    },
+    {
+      "epoch": 0.2661207778915046,
+      "grad_norm": 1.3841203451156616,
+      "learning_rate": 1.9877582492904533e-05,
+      "loss": 0.2377,
+      "step": 65
+    },
+    {
+      "epoch": 0.27021494370522003,
+      "grad_norm": 1.274598479270935,
+      "learning_rate": 1.9870502626379127e-05,
+      "loss": 0.2364,
+      "step": 66
+    },
+    {
+      "epoch": 0.2743091095189355,
+      "grad_norm": 1.59529447555542,
+      "learning_rate": 1.9863225079247286e-05,
+      "loss": 0.2647,
+      "step": 67
+    },
+    {
+      "epoch": 0.27840327533265097,
+      "grad_norm": 1.4986138343811035,
+      "learning_rate": 1.985574999725943e-05,
+      "loss": 0.264,
+      "step": 68
+    },
+    {
+      "epoch": 0.28249744114636643,
+      "grad_norm": 1.4692142009735107,
+      "learning_rate": 1.9848077530122083e-05,
+      "loss": 0.2475,
+      "step": 69
+    },
+    {
+      "epoch": 0.2865916069600819,
+      "grad_norm": 1.5602682828903198,
+      "learning_rate": 1.9840207831494903e-05,
+      "loss": 0.244,
+      "step": 70
+    },
+    {
+      "epoch": 0.2906857727737973,
+      "grad_norm": 1.5084633827209473,
+      "learning_rate": 1.983214105898757e-05,
+      "loss": 0.2442,
+      "step": 71
+    },
+    {
+      "epoch": 0.2947799385875128,
+      "grad_norm": 1.451074242591858,
+      "learning_rate": 1.9823877374156647e-05,
+      "loss": 0.2673,
+      "step": 72
+    },
+    {
+      "epoch": 0.29887410440122825,
+      "grad_norm": 1.4784746170043945,
+      "learning_rate": 1.9815416942502346e-05,
+      "loss": 0.2441,
+      "step": 73
+    },
+    {
+      "epoch": 0.3029682702149437,
+      "grad_norm": 1.368421196937561,
+      "learning_rate": 1.98067599334652e-05,
+      "loss": 0.2439,
+      "step": 74
+    },
+    {
+      "epoch": 0.3070624360286592,
+      "grad_norm": 1.4379913806915283,
+      "learning_rate": 1.979790652042268e-05,
+      "loss": 0.2238,
+      "step": 75
+    },
+    {
+      "epoch": 0.3111566018423746,
+      "grad_norm": 1.5823594331741333,
+      "learning_rate": 1.978885688068572e-05,
+      "loss": 0.2725,
+      "step": 76
+    },
+    {
+      "epoch": 0.31525076765609006,
+      "grad_norm": 1.5007753372192383,
+      "learning_rate": 1.9779611195495177e-05,
+      "loss": 0.2375,
+      "step": 77
+    },
+    {
+      "epoch": 0.3193449334698055,
+      "grad_norm": 1.473763346672058,
+      "learning_rate": 1.977016965001817e-05,
+      "loss": 0.2611,
+      "step": 78
+    },
+    {
+      "epoch": 0.323439099283521,
+      "grad_norm": 1.7358968257904053,
+      "learning_rate": 1.976053243334442e-05,
+      "loss": 0.2623,
+      "step": 79
+    },
+    {
+      "epoch": 0.32753326509723646,
+      "grad_norm": 1.5595204830169678,
+      "learning_rate": 1.9750699738482403e-05,
+      "loss": 0.2676,
+      "step": 80
+    },
+    {
+      "epoch": 0.33162743091095187,
+      "grad_norm": 1.685775637626648,
+      "learning_rate": 1.9740671762355548e-05,
+      "loss": 0.267,
+      "step": 81
+    },
+    {
+      "epoch": 0.33572159672466734,
+      "grad_norm": 1.3360930681228638,
+      "learning_rate": 1.973044870579824e-05,
+      "loss": 0.2339,
+      "step": 82
+    },
+    {
+      "epoch": 0.33572159672466734,
+      "eval_loss": 0.27538755536079407,
+      "eval_runtime": 5.5418,
+      "eval_samples_per_second": 14.255,
+      "eval_steps_per_second": 1.804,
+      "step": 82
+    },
+    {
+      "epoch": 0.3398157625383828,
+      "grad_norm": 1.7223079204559326,
+      "learning_rate": 1.972003077355183e-05,
+      "loss": 0.2843,
+      "step": 83
+    },
+    {
+      "epoch": 0.34390992835209827,
+      "grad_norm": 1.5389612913131714,
+      "learning_rate": 1.9709418174260523e-05,
+      "loss": 0.2604,
+      "step": 84
+    },
+    {
+      "epoch": 0.34800409416581374,
+      "grad_norm": 1.3990126848220825,
+      "learning_rate": 1.9698611120467196e-05,
+      "loss": 0.2588,
+      "step": 85
+    },
+    {
+      "epoch": 0.35209825997952915,
+      "grad_norm": 1.3090544939041138,
+      "learning_rate": 1.9687609828609156e-05,
+      "loss": 0.2264,
+      "step": 86
+    },
+    {
+      "epoch": 0.3561924257932446,
+      "grad_norm": 1.3547587394714355,
+      "learning_rate": 1.9676414519013782e-05,
+      "loss": 0.2436,
+      "step": 87
+    },
+    {
+      "epoch": 0.3602865916069601,
+      "grad_norm": 1.7848924398422241,
+      "learning_rate": 1.966502541589414e-05,
+      "loss": 0.2617,
+      "step": 88
+    },
+    {
+      "epoch": 0.36438075742067555,
+      "grad_norm": 1.450150728225708,
+      "learning_rate": 1.965344274734447e-05,
+      "loss": 0.2456,
+      "step": 89
+    },
+    {
+      "epoch": 0.368474923234391,
+      "grad_norm": 1.5955520868301392,
+      "learning_rate": 1.9641666745335626e-05,
+      "loss": 0.2571,
+      "step": 90
+    },
+    {
+      "epoch": 0.3725690890481064,
+      "grad_norm": 1.4538159370422363,
+      "learning_rate": 1.9629697645710432e-05,
+      "loss": 0.2552,
+      "step": 91
+    },
+    {
+      "epoch": 0.3766632548618219,
+      "grad_norm": 1.4258908033370972,
+      "learning_rate": 1.961753568817896e-05,
+      "loss": 0.2402,
+      "step": 92
+    },
+    {
+      "epoch": 0.38075742067553736,
+      "grad_norm": 1.463593602180481,
+      "learning_rate": 1.9605181116313725e-05,
+      "loss": 0.2612,
+      "step": 93
+    },
+    {
+      "epoch": 0.38485158648925283,
+      "grad_norm": 1.4867630004882812,
+      "learning_rate": 1.9592634177544803e-05,
+      "loss": 0.2272,
+      "step": 94
+    },
+    {
+      "epoch": 0.3889457523029683,
+      "grad_norm": 1.4335721731185913,
+      "learning_rate": 1.957989512315489e-05,
+      "loss": 0.2562,
+      "step": 95
+    },
+    {
+      "epoch": 0.3930399181166837,
+      "grad_norm": 1.3976503610610962,
+      "learning_rate": 1.9566964208274254e-05,
+      "loss": 0.2527,
+      "step": 96
+    },
+    {
+      "epoch": 0.3971340839303992,
+      "grad_norm": 1.455560564994812,
+      "learning_rate": 1.9553841691875632e-05,
+      "loss": 0.2617,
+      "step": 97
+    },
+    {
+      "epoch": 0.40122824974411464,
+      "grad_norm": 1.319207787513733,
+      "learning_rate": 1.9540527836769047e-05,
+      "loss": 0.251,
+      "step": 98
+    },
+    {
+      "epoch": 0.4053224155578301,
+      "grad_norm": 1.3503410816192627,
+      "learning_rate": 1.9527022909596537e-05,
+      "loss": 0.253,
+      "step": 99
+    },
+    {
+      "epoch": 0.4094165813715456,
+      "grad_norm": 1.472091555595398,
+      "learning_rate": 1.951332718082682e-05,
+      "loss": 0.2935,
+      "step": 100
+    },
+    {
+      "epoch": 0.413510747185261,
+      "grad_norm": 1.4811311960220337,
+      "learning_rate": 1.9499440924749878e-05,
+      "loss": 0.261,
+      "step": 101
+    },
+    {
+      "epoch": 0.41760491299897645,
+      "grad_norm": 1.595140814781189,
+      "learning_rate": 1.9485364419471454e-05,
+      "loss": 0.2703,
+      "step": 102
+    },
+    {
+      "epoch": 0.4216990788126919,
+      "grad_norm": 1.3853557109832764,
+      "learning_rate": 1.9471097946907506e-05,
+      "loss": 0.2345,
+      "step": 103
+    },
+    {
+      "epoch": 0.4257932446264074,
+      "grad_norm": 1.5886335372924805,
+      "learning_rate": 1.9456641792778527e-05,
+      "loss": 0.269,
+      "step": 104
+    },
+    {
+      "epoch": 0.42988741044012285,
+      "grad_norm": 1.3403784036636353,
+      "learning_rate": 1.9441996246603848e-05,
+      "loss": 0.234,
+      "step": 105
+    },
+    {
+      "epoch": 0.43398157625383826,
+      "grad_norm": 1.4540488719940186,
+      "learning_rate": 1.9427161601695833e-05,
+      "loss": 0.246,
+      "step": 106
+    },
+    {
+      "epoch": 0.43807574206755373,
+      "grad_norm": 1.5493131875991821,
+      "learning_rate": 1.9412138155154e-05,
+      "loss": 0.2284,
+      "step": 107
+    },
+    {
+      "epoch": 0.4421699078812692,
+      "grad_norm": 1.290971279144287,
+      "learning_rate": 1.9396926207859085e-05,
+      "loss": 0.2392,
+      "step": 108
+    },
+    {
+      "epoch": 0.44626407369498466,
+      "grad_norm": 1.4140104055404663,
+      "learning_rate": 1.9381526064466995e-05,
+      "loss": 0.2554,
+      "step": 109
+    },
+    {
+      "epoch": 0.4503582395087001,
+      "grad_norm": 1.7200373411178589,
+      "learning_rate": 1.9365938033402715e-05,
+      "loss": 0.2359,
+      "step": 110
+    },
+    {
+      "epoch": 0.45445240532241554,
+      "grad_norm": 1.2797805070877075,
+      "learning_rate": 1.9350162426854152e-05,
+      "loss": 0.2321,
+      "step": 111
+    },
+    {
+      "epoch": 0.458546571136131,
+      "grad_norm": 1.326955795288086,
+      "learning_rate": 1.933419956076584e-05,
+      "loss": 0.2516,
+      "step": 112
+    },
+    {
+      "epoch": 0.4626407369498465,
+      "grad_norm": 1.510201334953308,
+      "learning_rate": 1.9318049754832656e-05,
+      "loss": 0.2467,
+      "step": 113
+    },
+    {
+      "epoch": 0.46673490276356194,
+      "grad_norm": 2.9965062141418457,
+      "learning_rate": 1.9301713332493386e-05,
+      "loss": 0.2587,
+      "step": 114
+    },
+    {
+      "epoch": 0.47082906857727735,
+      "grad_norm": 1.3663560152053833,
+      "learning_rate": 1.9285190620924267e-05,
+      "loss": 0.2535,
+      "step": 115
+    },
+    {
+      "epoch": 0.4749232343909928,
+      "grad_norm": 1.1917448043823242,
+      "learning_rate": 1.926848195103242e-05,
+      "loss": 0.2292,
+      "step": 116
+    },
+    {
+      "epoch": 0.4790174002047083,
+      "grad_norm": 1.3093336820602417,
+      "learning_rate": 1.925158765744924e-05,
+      "loss": 0.229,
+      "step": 117
+    },
+    {
+      "epoch": 0.48311156601842375,
+      "grad_norm": 1.5121235847473145,
+      "learning_rate": 1.923450807852367e-05,
+      "loss": 0.2527,
+      "step": 118
+    },
+    {
+      "epoch": 0.4872057318321392,
+      "grad_norm": 1.461378812789917,
+      "learning_rate": 1.9217243556315445e-05,
+      "loss": 0.2631,
+      "step": 119
+    },
+    {
+      "epoch": 0.49129989764585463,
+      "grad_norm": 1.4177104234695435,
+      "learning_rate": 1.9199794436588244e-05,
+      "loss": 0.2378,
+      "step": 120
+    },
+    {
+      "epoch": 0.4953940634595701,
+      "grad_norm": 1.5456838607788086,
+      "learning_rate": 1.9182161068802742e-05,
+      "loss": 0.2491,
+      "step": 121
+    },
+    {
+      "epoch": 0.49948822927328557,
+      "grad_norm": 1.554958701133728,
+      "learning_rate": 1.916434380610963e-05,
+      "loss": 0.2461,
+      "step": 122
+    },
+    {
+      "epoch": 0.503582395087001,
+      "grad_norm": 1.4894706010818481,
+      "learning_rate": 1.9146343005342546e-05,
+      "loss": 0.2823,
+      "step": 123
+    },
+    {
+      "epoch": 0.5076765609007164,
+      "grad_norm": 1.359287142753601,
+      "learning_rate": 1.912815902701091e-05,
+      "loss": 0.2366,
+      "step": 124
+    },
+    {
+      "epoch": 0.5117707267144319,
+      "grad_norm": 1.4244722127914429,
+      "learning_rate": 1.9109792235292715e-05,
+      "loss": 0.2398,
+      "step": 125
+    },
+    {
+      "epoch": 0.5158648925281474,
+      "grad_norm": 1.4032812118530273,
+      "learning_rate": 1.909124299802724e-05,
+      "loss": 0.2601,
+      "step": 126
+    },
+    {
+      "epoch": 0.5199590583418628,
+      "grad_norm": 1.5199010372161865,
+      "learning_rate": 1.9072511686707663e-05,
+      "loss": 0.2458,
+      "step": 127
+    },
+    {
+      "epoch": 0.5240532241555783,
+      "grad_norm": 1.4342416524887085,
+      "learning_rate": 1.9053598676473656e-05,
+      "loss": 0.241,
+      "step": 128
+    },
+    {
+      "epoch": 0.5281473899692938,
+      "grad_norm": 1.338181972503662,
+      "learning_rate": 1.9034504346103825e-05,
+      "loss": 0.2508,
+      "step": 129
+    },
+    {
+      "epoch": 0.5322415557830092,
+      "grad_norm": 1.492775559425354,
+      "learning_rate": 1.9015229078008163e-05,
+      "loss": 0.2626,
+      "step": 130
+    },
+    {
+      "epoch": 0.5363357215967247,
+      "grad_norm": 1.3123077154159546,
+      "learning_rate": 1.8995773258220374e-05,
+      "loss": 0.258,
+      "step": 131
+    },
+    {
+      "epoch": 0.5404298874104401,
+      "grad_norm": 1.4001896381378174,
+      "learning_rate": 1.8976137276390145e-05,
+      "loss": 0.2425,
+      "step": 132
+    },
+    {
+      "epoch": 0.5445240532241555,
+      "grad_norm": 1.3989671468734741,
+      "learning_rate": 1.8956321525775337e-05,
+      "loss": 0.2781,
+      "step": 133
+    },
+    {
+      "epoch": 0.548618219037871,
+      "grad_norm": 1.260289192199707,
+      "learning_rate": 1.8936326403234125e-05,
+      "loss": 0.2432,
+      "step": 134
+    },
+    {
+      "epoch": 0.5527123848515865,
+      "grad_norm": 1.308370590209961,
+      "learning_rate": 1.891615230921703e-05,
+      "loss": 0.2408,
+      "step": 135
+    },
+    {
+      "epoch": 0.5568065506653019,
+      "grad_norm": 1.2668206691741943,
+      "learning_rate": 1.8895799647758912e-05,
+      "loss": 0.2408,
+      "step": 136
+    },
+    {
+      "epoch": 0.5609007164790174,
+      "grad_norm": 1.4105634689331055,
+      "learning_rate": 1.8875268826470875e-05,
+      "loss": 0.2688,
+      "step": 137
+    },
+    {
+      "epoch": 0.5649948822927329,
+      "grad_norm": 1.3877664804458618,
+      "learning_rate": 1.8854560256532098e-05,
+      "loss": 0.2379,
+      "step": 138
+    },
+    {
+      "epoch": 0.5690890481064483,
+      "grad_norm": 1.2643476724624634,
+      "learning_rate": 1.8833674352681613e-05,
+      "loss": 0.2375,
+      "step": 139
+    },
+    {
+      "epoch": 0.5731832139201638,
+      "grad_norm": 1.6310402154922485,
+      "learning_rate": 1.881261153320999e-05,
+      "loss": 0.2435,
+      "step": 140
+    },
+    {
+      "epoch": 0.5772773797338793,
+      "grad_norm": 1.35072660446167,
+      "learning_rate": 1.879137221995095e-05,
+      "loss": 0.2293,
+      "step": 141
+    },
+    {
+      "epoch": 0.5813715455475946,
+      "grad_norm": 1.409683346748352,
+      "learning_rate": 1.8769956838272937e-05,
+      "loss": 0.2584,
+      "step": 142
+    },
+    {
+      "epoch": 0.5854657113613101,
+      "grad_norm": 1.4319274425506592,
+      "learning_rate": 1.8748365817070586e-05,
+      "loss": 0.2669,
+      "step": 143
+    },
+    {
+      "epoch": 0.5895598771750256,
+      "grad_norm": 1.4259189367294312,
+      "learning_rate": 1.8726599588756144e-05,
+      "loss": 0.2447,
+      "step": 144
+    },
+    {
+      "epoch": 0.593654042988741,
+      "grad_norm": 1.4267385005950928,
+      "learning_rate": 1.8704658589250795e-05,
+      "loss": 0.2567,
+      "step": 145
+    },
+    {
+      "epoch": 0.5977482088024565,
+      "grad_norm": 1.4896032810211182,
+      "learning_rate": 1.868254325797594e-05,
+      "loss": 0.2363,
+      "step": 146
+    },
+    {
+      "epoch": 0.601842374616172,
+      "grad_norm": 1.355659008026123,
+      "learning_rate": 1.866025403784439e-05,
+      "loss": 0.2481,
+      "step": 147
+    },
+    {
+      "epoch": 0.6059365404298874,
+      "grad_norm": 1.2849416732788086,
+      "learning_rate": 1.8637791375251505e-05,
+      "loss": 0.2631,
+      "step": 148
+    },
+    {
+      "epoch": 0.6100307062436029,
+      "grad_norm": 1.279733419418335,
+      "learning_rate": 1.8615155720066247e-05,
+      "loss": 0.2487,
+      "step": 149
+    },
+    {
+      "epoch": 0.6141248720573184,
+      "grad_norm": 1.370506763458252,
+      "learning_rate": 1.859234752562217e-05,
+      "loss": 0.245,
+      "step": 150
+    },
+    {
+      "epoch": 0.6182190378710338,
+      "grad_norm": 1.352271556854248,
+      "learning_rate": 1.8569367248708343e-05,
+      "loss": 0.227,
+      "step": 151
+    },
+    {
+      "epoch": 0.6223132036847492,
+      "grad_norm": 1.3176748752593994,
+      "learning_rate": 1.8546215349560204e-05,
+      "loss": 0.2484,
+      "step": 152
+    },
+    {
+      "epoch": 0.6264073694984647,
+      "grad_norm": 1.386770486831665,
+      "learning_rate": 1.8522892291850335e-05,
+      "loss": 0.257,
+      "step": 153
+    },
+    {
+      "epoch": 0.6305015353121801,
+      "grad_norm": 1.5251948833465576,
+      "learning_rate": 1.849939854267919e-05,
+      "loss": 0.2727,
+      "step": 154
+    },
+    {
+      "epoch": 0.6345957011258956,
+      "grad_norm": 1.378936767578125,
+      "learning_rate": 1.847573457256571e-05,
+      "loss": 0.2337,
+      "step": 155
+    },
+    {
+      "epoch": 0.638689866939611,
+      "grad_norm": 1.3500293493270874,
+      "learning_rate": 1.845190085543795e-05,
+      "loss": 0.2618,
+      "step": 156
+    },
+    {
+      "epoch": 0.6427840327533265,
+      "grad_norm": 1.4462950229644775,
+      "learning_rate": 1.8427897868623535e-05,
+      "loss": 0.2588,
+      "step": 157
+    },
+    {
+      "epoch": 0.646878198567042,
+      "grad_norm": 1.2208290100097656,
+      "learning_rate": 1.840372609284013e-05,
+      "loss": 0.2927,
+      "step": 158
+    },
+    {
+      "epoch": 0.6509723643807575,
+      "grad_norm": 1.3487396240234375,
+      "learning_rate": 1.8379386012185813e-05,
+      "loss": 0.2417,
+      "step": 159
+    },
+    {
+      "epoch": 0.6550665301944729,
+      "grad_norm": 1.2655378580093384,
+      "learning_rate": 1.8354878114129368e-05,
+      "loss": 0.2428,
+      "step": 160
+    },
+    {
+      "epoch": 0.6591606960081884,
+      "grad_norm": 1.181028962135315,
+      "learning_rate": 1.8330202889500518e-05,
+      "loss": 0.2397,
+      "step": 161
+    },
+    {
+      "epoch": 0.6632548618219037,
+      "grad_norm": 1.3897309303283691,
+      "learning_rate": 1.8305360832480118e-05,
+      "loss": 0.2276,
+      "step": 162
+    },
+    {
+      "epoch": 0.6673490276356192,
+      "grad_norm": 1.340640902519226,
+      "learning_rate": 1.8280352440590236e-05,
+      "loss": 0.2375,
+      "step": 163
+    },
+    {
+      "epoch": 0.6714431934493347,
+      "grad_norm": 1.5388420820236206,
+      "learning_rate": 1.82551782146842e-05,
+      "loss": 0.2318,
+      "step": 164
+    },
+    {
+      "epoch": 0.6714431934493347,
+      "eval_loss": 0.2667195796966553,
+      "eval_runtime": 6.096,
+      "eval_samples_per_second": 12.959,
+      "eval_steps_per_second": 1.64,
+      "step": 164
+    },
+    {
+      "epoch": 0.6755373592630501,
+      "grad_norm": 1.4711631536483765,
+      "learning_rate": 1.8229838658936566e-05,
+      "loss": 0.2367,
+      "step": 165
+    },
+    {
+      "epoch": 0.6796315250767656,
+      "grad_norm": 1.3367588520050049,
+      "learning_rate": 1.8204334280833005e-05,
+      "loss": 0.2652,
+      "step": 166
+    },
+    {
+      "epoch": 0.6837256908904811,
+      "grad_norm": 1.5473802089691162,
+      "learning_rate": 1.817866559116017e-05,
+      "loss": 0.2693,
+      "step": 167
+    },
+    {
+      "epoch": 0.6878198567041965,
+      "grad_norm": 1.3539572954177856,
+      "learning_rate": 1.8152833103995443e-05,
+      "loss": 0.2431,
+      "step": 168
+    },
+    {
+      "epoch": 0.691914022517912,
+      "grad_norm": 1.3012574911117554,
+      "learning_rate": 1.8126837336696645e-05,
+      "loss": 0.2602,
+      "step": 169
+    },
+    {
+      "epoch": 0.6960081883316275,
+      "grad_norm": 1.3125128746032715,
+      "learning_rate": 1.8100678809891668e-05,
+      "loss": 0.2426,
+      "step": 170
+    },
+    {
+      "epoch": 0.7001023541453428,
+      "grad_norm": 1.2353239059448242,
+      "learning_rate": 1.807435804746807e-05,
+      "loss": 0.2176,
+      "step": 171
+    },
+    {
+      "epoch": 0.7041965199590583,
+      "grad_norm": 1.268622636795044,
+      "learning_rate": 1.8047875576562556e-05,
+      "loss": 0.2495,
+      "step": 172
+    },
+    {
+      "epoch": 0.7082906857727738,
+      "grad_norm": 1.3845950365066528,
+      "learning_rate": 1.802123192755044e-05,
+      "loss": 0.2643,
+      "step": 173
+    },
+    {
+      "epoch": 0.7123848515864892,
+      "grad_norm": 1.2450510263442993,
+      "learning_rate": 1.7994427634035016e-05,
+      "loss": 0.2445,
+      "step": 174
+    },
+    {
+      "epoch": 0.7164790174002047,
+      "grad_norm": 1.545483946800232,
+      "learning_rate": 1.796746323283686e-05,
+      "loss": 0.2594,
+      "step": 175
+    },
+    {
+      "epoch": 0.7205731832139202,
+      "grad_norm": 1.3044480085372925,
+      "learning_rate": 1.7940339263983112e-05,
+      "loss": 0.2563,
+      "step": 176
+    },
+    {
+      "epoch": 0.7246673490276356,
+      "grad_norm": 1.288320779800415,
+      "learning_rate": 1.791305627069662e-05,
+      "loss": 0.2478,
+      "step": 177
+    },
+    {
+      "epoch": 0.7287615148413511,
+      "grad_norm": 1.2802083492279053,
+      "learning_rate": 1.7885614799385086e-05,
+      "loss": 0.2331,
+      "step": 178
+    },
+    {
+      "epoch": 0.7328556806550666,
+      "grad_norm": 1.3979520797729492,
+      "learning_rate": 1.785801539963012e-05,
+      "loss": 0.2458,
+      "step": 179
+    },
+    {
+      "epoch": 0.736949846468782,
+      "grad_norm": 1.2247169017791748,
+      "learning_rate": 1.7830258624176224e-05,
+      "loss": 0.2132,
+      "step": 180
+    },
+    {
+      "epoch": 0.7410440122824974,
+      "grad_norm": 1.4145004749298096,
+      "learning_rate": 1.7802345028919728e-05,
+      "loss": 0.2402,
+      "step": 181
+    },
+    {
+      "epoch": 0.7451381780962129,
+      "grad_norm": 1.5720893144607544,
+      "learning_rate": 1.777427517289766e-05,
+      "loss": 0.258,
+      "step": 182
+    },
+    {
+      "epoch": 0.7492323439099283,
+      "grad_norm": 1.272928237915039,
+      "learning_rate": 1.7746049618276545e-05,
+      "loss": 0.2501,
+      "step": 183
+    },
+    {
+      "epoch": 0.7533265097236438,
+      "grad_norm": 1.4003905057907104,
+      "learning_rate": 1.7717668930341152e-05,
+      "loss": 0.2567,
+      "step": 184
+    },
+    {
+      "epoch": 0.7574206755373593,
+      "grad_norm": 1.4310050010681152,
+      "learning_rate": 1.768913367748316e-05,
+      "loss": 0.2443,
+      "step": 185
+    },
+    {
+      "epoch": 0.7615148413510747,
+      "grad_norm": 1.480718731880188,
+      "learning_rate": 1.766044443118978e-05,
+      "loss": 0.2564,
+      "step": 186
+    },
+    {
+      "epoch": 0.7656090071647902,
+      "grad_norm": 1.331586480140686,
+      "learning_rate": 1.7631601766032337e-05,
+      "loss": 0.2467,
+      "step": 187
+    },
+    {
+      "epoch": 0.7697031729785057,
+      "grad_norm": 1.2713260650634766,
+      "learning_rate": 1.7602606259654704e-05,
+      "loss": 0.2406,
+      "step": 188
+    },
+    {
+      "epoch": 0.7737973387922211,
+      "grad_norm": 1.4929317235946655,
+      "learning_rate": 1.7573458492761802e-05,
+      "loss": 0.2515,
+      "step": 189
+    },
+    {
+      "epoch": 0.7778915046059366,
+      "grad_norm": 1.117374062538147,
+      "learning_rate": 1.7544159049107902e-05,
+      "loss": 0.2165,
+      "step": 190
+    },
+    {
+      "epoch": 0.781985670419652,
+      "grad_norm": 1.2868603467941284,
+      "learning_rate": 1.7514708515485002e-05,
+      "loss": 0.2469,
+      "step": 191
+    },
+    {
+      "epoch": 0.7860798362333674,
+      "grad_norm": 1.3611894845962524,
+      "learning_rate": 1.7485107481711014e-05,
+      "loss": 0.2468,
+      "step": 192
+    },
+    {
+      "epoch": 0.7901740020470829,
+      "grad_norm": 1.3682395219802856,
+      "learning_rate": 1.7455356540617988e-05,
+      "loss": 0.2548,
+      "step": 193
+    },
+    {
+      "epoch": 0.7942681678607983,
+      "grad_norm": 1.2504215240478516,
+      "learning_rate": 1.7425456288040236e-05,
+      "loss": 0.2563,
+      "step": 194
+    },
+    {
+      "epoch": 0.7983623336745138,
+      "grad_norm": 1.3082585334777832,
+      "learning_rate": 1.7395407322802374e-05,
+      "loss": 0.232,
+      "step": 195
+    },
+    {
+      "epoch": 0.8024564994882293,
+      "grad_norm": 1.409508466720581,
+      "learning_rate": 1.736521024670737e-05,
+      "loss": 0.2438,
+      "step": 196
+    },
+    {
+      "epoch": 0.8065506653019447,
+      "grad_norm": 1.1714893579483032,
+      "learning_rate": 1.733486566452446e-05,
+      "loss": 0.2201,
+      "step": 197
+    },
+    {
+      "epoch": 0.8106448311156602,
+      "grad_norm": 1.3167288303375244,
+      "learning_rate": 1.7304374183977032e-05,
+      "loss": 0.2359,
+      "step": 198
+    },
+    {
+      "epoch": 0.8147389969293757,
+      "grad_norm": 1.2892719507217407,
+      "learning_rate": 1.7273736415730488e-05,
+      "loss": 0.2373,
+      "step": 199
+    },
+    {
+      "epoch": 0.8188331627430911,
+      "grad_norm": 1.2318534851074219,
+      "learning_rate": 1.7242952973379983e-05,
+      "loss": 0.2493,
+      "step": 200
+    },
+    {
+      "epoch": 0.8229273285568065,
+      "grad_norm": 1.2666089534759521,
+      "learning_rate": 1.7212024473438145e-05,
+      "loss": 0.2335,
+      "step": 201
+    },
+    {
+      "epoch": 0.827021494370522,
+      "grad_norm": 1.3021701574325562,
+      "learning_rate": 1.7180951535322742e-05,
+      "loss": 0.2457,
+      "step": 202
+    },
+    {
+      "epoch": 0.8311156601842374,
+      "grad_norm": 1.2390035390853882,
+      "learning_rate": 1.7149734781344247e-05,
+      "loss": 0.2317,
+      "step": 203
+    },
+    {
+      "epoch": 0.8352098259979529,
+      "grad_norm": 1.373651146888733,
+      "learning_rate": 1.7118374836693407e-05,
+      "loss": 0.2477,
+      "step": 204
+    },
+    {
+      "epoch": 0.8393039918116684,
+      "grad_norm": 1.4125158786773682,
+      "learning_rate": 1.7086872329428702e-05,
+      "loss": 0.2716,
+      "step": 205
+    },
+    {
+      "epoch": 0.8433981576253838,
+      "grad_norm": 1.3470803499221802,
+      "learning_rate": 1.705522789046377e-05,
+      "loss": 0.264,
+      "step": 206
+    },
+    {
+      "epoch": 0.8474923234390993,
+      "grad_norm": 1.419498324394226,
+      "learning_rate": 1.7023442153554776e-05,
+      "loss": 0.2626,
+      "step": 207
+    },
+    {
+      "epoch": 0.8515864892528148,
+      "grad_norm": 1.4023360013961792,
+      "learning_rate": 1.6991515755287715e-05,
+      "loss": 0.2786,
+      "step": 208
+    },
+    {
+      "epoch": 0.8556806550665302,
+      "grad_norm": 1.3539841175079346,
+      "learning_rate": 1.695944933506567e-05,
+      "loss": 0.2623,
+      "step": 209
+    },
+    {
+      "epoch": 0.8597748208802457,
+      "grad_norm": 1.4278448820114136,
+      "learning_rate": 1.6927243535095995e-05,
+      "loss": 0.2491,
+      "step": 210
+    },
+    {
+      "epoch": 0.8638689866939611,
+      "grad_norm": 1.4281779527664185,
+      "learning_rate": 1.6894899000377462e-05,
+      "loss": 0.2303,
+      "step": 211
+    },
+    {
+      "epoch": 0.8679631525076765,
+      "grad_norm": 1.2435628175735474,
+      "learning_rate": 1.686241637868734e-05,
+      "loss": 0.254,
+      "step": 212
+    },
+    {
+      "epoch": 0.872057318321392,
+      "grad_norm": 1.2799077033996582,
+      "learning_rate": 1.6829796320568416e-05,
+      "loss": 0.2253,
+      "step": 213
+    },
+    {
+      "epoch": 0.8761514841351075,
+      "grad_norm": 1.4002091884613037,
+      "learning_rate": 1.6797039479315994e-05,
+      "loss": 0.2568,
+      "step": 214
+    },
+    {
+      "epoch": 0.8802456499488229,
+      "grad_norm": 1.2796311378479004,
+      "learning_rate": 1.6764146510964762e-05,
+      "loss": 0.2488,
+      "step": 215
+    },
+    {
+      "epoch": 0.8843398157625384,
+      "grad_norm": 1.28677237033844,
+      "learning_rate": 1.67311180742757e-05,
+      "loss": 0.2619,
+      "step": 216
+    },
+    {
+      "epoch": 0.8884339815762539,
+      "grad_norm": 1.5529658794403076,
+      "learning_rate": 1.669795483072287e-05,
+      "loss": 0.2312,
+      "step": 217
+    },
+    {
+      "epoch": 0.8925281473899693,
+      "grad_norm": 1.2904109954833984,
+      "learning_rate": 1.6664657444480145e-05,
+      "loss": 0.2366,
+      "step": 218
+    },
+    {
+      "epoch": 0.8966223132036848,
+      "grad_norm": 1.2689939737319946,
+      "learning_rate": 1.6631226582407954e-05,
+      "loss": 0.2446,
+      "step": 219
+    },
+    {
+      "epoch": 0.9007164790174002,
+      "grad_norm": 1.2608931064605713,
+      "learning_rate": 1.6597662914039885e-05,
+      "loss": 0.2419,
+      "step": 220
+    },
+    {
+      "epoch": 0.9048106448311156,
+      "grad_norm": 1.2563594579696655,
+      "learning_rate": 1.65639671115693e-05,
+      "loss": 0.2363,
+      "step": 221
+    },
+    {
+      "epoch": 0.9089048106448311,
+      "grad_norm": 1.320057988166809,
+      "learning_rate": 1.653013984983585e-05,
+      "loss": 0.2508,
+      "step": 222
+    },
+    {
+      "epoch": 0.9129989764585466,
+      "grad_norm": 1.2236367464065552,
+      "learning_rate": 1.6496181806312005e-05,
+      "loss": 0.2263,
+      "step": 223
+    },
+    {
+      "epoch": 0.917093142272262,
+      "grad_norm": 1.293455958366394,
+      "learning_rate": 1.6462093661089432e-05,
+      "loss": 0.2504,
+      "step": 224
+    },
+    {
+      "epoch": 0.9211873080859775,
+      "grad_norm": 1.3519924879074097,
+      "learning_rate": 1.6427876096865394e-05,
+      "loss": 0.2458,
+      "step": 225
+    },
+    {
+      "epoch": 0.925281473899693,
+      "grad_norm": 1.1655514240264893,
+      "learning_rate": 1.6393529798929103e-05,
+      "loss": 0.249,
+      "step": 226
+    },
+    {
+      "epoch": 0.9293756397134084,
+      "grad_norm": 1.3402174711227417,
+      "learning_rate": 1.635905545514795e-05,
+      "loss": 0.2426,
+      "step": 227
+    },
+    {
+      "epoch": 0.9334698055271239,
+      "grad_norm": 1.3474458456039429,
+      "learning_rate": 1.6324453755953772e-05,
+      "loss": 0.2497,
+      "step": 228
+    },
+    {
+      "epoch": 0.9375639713408394,
+      "grad_norm": 1.2284787893295288,
+      "learning_rate": 1.6289725394328998e-05,
+      "loss": 0.2438,
+      "step": 229
+    },
+    {
+      "epoch": 0.9416581371545547,
+      "grad_norm": 1.2306231260299683,
+      "learning_rate": 1.6254871065792776e-05,
+      "loss": 0.248,
+      "step": 230
+    },
+    {
+      "epoch": 0.9457523029682702,
+      "grad_norm": 1.3834123611450195,
+      "learning_rate": 1.621989146838704e-05,
+      "loss": 0.2798,
+      "step": 231
+    },
+    {
+      "epoch": 0.9498464687819856,
+      "grad_norm": 1.333827257156372,
+      "learning_rate": 1.618478730266255e-05,
+      "loss": 0.2585,
+      "step": 232
+    },
+    {
+      "epoch": 0.9539406345957011,
+      "grad_norm": 1.272150993347168,
+      "learning_rate": 1.6149559271664835e-05,
+      "loss": 0.2512,
+      "step": 233
+    },
+    {
+      "epoch": 0.9580348004094166,
+      "grad_norm": 1.3800173997879028,
+      "learning_rate": 1.6114208080920125e-05,
+      "loss": 0.2483,
+      "step": 234
+    },
+    {
+      "epoch": 0.962128966223132,
+      "grad_norm": 1.2623674869537354,
+      "learning_rate": 1.607873443842122e-05,
+      "loss": 0.2447,
+      "step": 235
+    },
+    {
+      "epoch": 0.9662231320368475,
+      "grad_norm": 1.2797845602035522,
+      "learning_rate": 1.6043139054613326e-05,
+      "loss": 0.2258,
+      "step": 236
+    },
+    {
+      "epoch": 0.970317297850563,
+      "grad_norm": 1.3689030408859253,
+      "learning_rate": 1.600742264237979e-05,
+      "loss": 0.247,
+      "step": 237
+    },
+    {
+      "epoch": 0.9744114636642784,
+      "grad_norm": 1.3384076356887817,
+      "learning_rate": 1.5971585917027864e-05,
+      "loss": 0.2286,
+      "step": 238
+    },
+    {
+      "epoch": 0.9785056294779939,
+      "grad_norm": 1.210942268371582,
+      "learning_rate": 1.5935629596274345e-05,
+      "loss": 0.2402,
+      "step": 239
+    },
+    {
+      "epoch": 0.9825997952917093,
+      "grad_norm": 1.2547039985656738,
+      "learning_rate": 1.5899554400231233e-05,
+      "loss": 0.2386,
+      "step": 240
+    },
+    {
+      "epoch": 0.9866939611054247,
+      "grad_norm": 1.337138056755066,
+      "learning_rate": 1.586336105139127e-05,
+      "loss": 0.2686,
+      "step": 241
+    },
+    {
+      "epoch": 0.9907881269191402,
+      "grad_norm": 1.273237943649292,
+      "learning_rate": 1.5827050274613512e-05,
+      "loss": 0.2653,
+      "step": 242
+    },
+    {
+      "epoch": 0.9948822927328557,
+      "grad_norm": 1.3556797504425049,
+      "learning_rate": 1.579062279710879e-05,
+      "loss": 0.2365,
+      "step": 243
+    },
+    {
+      "epoch": 0.9989764585465711,
+      "grad_norm": 1.2234649658203125,
+      "learning_rate": 1.5754079348425137e-05,
+      "loss": 0.239,
+      "step": 244
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 732,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 244,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.94188710905643e+17,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}

3b-w-cot+/checkpoint-244/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:632e641a15180cc852702434a272df94b8012efb84c5e296eb59b1554cdab170
+size 10744

3b-w-cot+/checkpoint-244/zero_to_fp32.py

@@ -0,0 +1,760 @@
+#!/usr/bin/env python
+
+# Copyright (c) Microsoft Corporation.
+# SPDX-License-Identifier: Apache-2.0
+
+# DeepSpeed Team
+
+# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets
+# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in
+# the future. Once extracted, the weights don't require DeepSpeed and can be used in any
+# application.
+#
+# example:
+#   python zero_to_fp32.py . output_dir/
+#   or
+#   python zero_to_fp32.py . output_dir/ --safe_serialization
+
+import argparse
+import torch
+import glob
+import math
+import os
+import re
+import gc
+import json
+import numpy as np
+from tqdm import tqdm
+from collections import OrderedDict
+from dataclasses import dataclass
+
+# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with
+# DeepSpeed data structures it has to be available in the current python environment.
+from deepspeed.utils import logger
+from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS,
+                                            FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES,
+                                            FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS)
+
+
+@dataclass
+class zero_model_state:
+    buffers: dict()
+    param_shapes: dict()
+    shared_params: list
+    ds_version: int
+    frozen_param_shapes: dict()
+    frozen_param_fragments: dict()
+
+
+debug = 0
+
+# load to cpu
+device = torch.device('cpu')
+
+
+def atoi(text):
+    return int(text) if text.isdigit() else text
+
+
+def natural_keys(text):
+    '''
+    alist.sort(key=natural_keys) sorts in human order
+    http://nedbatchelder.com/blog/200712/human_sorting.html
+    (See Toothy's implementation in the comments)
+    '''
+    return [atoi(c) for c in re.split(r'(\d+)', text)]
+
+
+def get_model_state_file(checkpoint_dir, zero_stage):
+    if not os.path.isdir(checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist")
+
+    # there should be only one file
+    if zero_stage <= 2:
+        file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt")
+    elif zero_stage == 3:
+        file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt")
+
+    if not os.path.exists(file):
+        raise FileNotFoundError(f"can't find model states file at '{file}'")
+
+    return file
+
+
+def get_checkpoint_files(checkpoint_dir, glob_pattern):
+    # XXX: need to test that this simple glob rule works for multi-node setup too
+    ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys)
+
+    if len(ckpt_files) == 0:
+        raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'")
+
+    return ckpt_files
+
+
+def get_optim_files(checkpoint_dir):
+    return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt")
+
+
+def get_model_state_files(checkpoint_dir):
+    return get_checkpoint_files(checkpoint_dir, "*_model_states.pt")
+
+
+def parse_model_states(files):
+    zero_model_states = []
+    for file in files:
+        state_dict = torch.load(file, map_location=device, weights_only=False)
+
+        if BUFFER_NAMES not in state_dict:
+            raise ValueError(f"{file} is not a model state checkpoint")
+        buffer_names = state_dict[BUFFER_NAMES]
+        if debug:
+            print("Found buffers:", buffer_names)
+
+        # recover just the buffers while restoring them to fp32 if they were saved in fp16
+        buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names}
+        param_shapes = state_dict[PARAM_SHAPES]
+
+        # collect parameters that are included in param_shapes
+        param_names = []
+        for s in param_shapes:
+            for name in s.keys():
+                param_names.append(name)
+
+        # update with frozen parameters
+        frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None)
+        if frozen_param_shapes is not None:
+            if debug:
+                print(f"Found frozen_param_shapes: {frozen_param_shapes}")
+            param_names += list(frozen_param_shapes.keys())
+
+        # handle shared params
+        shared_params = [[k, v] for k, v in state_dict["shared_params"].items()]
+
+        ds_version = state_dict.get(DS_VERSION, None)
+
+        frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None)
+
+        z_model_state = zero_model_state(buffers=buffers,
+                                         param_shapes=param_shapes,
+                                         shared_params=shared_params,
+                                         ds_version=ds_version,
+                                         frozen_param_shapes=frozen_param_shapes,
+                                         frozen_param_fragments=frozen_param_fragments)
+        zero_model_states.append(z_model_state)
+
+    return zero_model_states
+
+
+def parse_optim_states(files, ds_checkpoint_dir):
+    total_files = len(files)
+    state_dicts = []
+    for f in tqdm(files, desc='Loading checkpoint shards'):
+        state_dict = torch.load(f, map_location=device, mmap=True, weights_only=False)
+        # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights
+        # and also handle the case where it was already removed by another helper script
+        state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None)
+        state_dicts.append(state_dict)
+
+    if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]:
+        raise ValueError(f"{files[0]} is not a zero checkpoint")
+    zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE]
+    world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT]
+
+    # For ZeRO-2 each param group can have different partition_count as data parallelism for expert
+    # parameters can be different from data parallelism for non-expert parameters. So we can just
+    # use the max of the partition_count to get the dp world_size.
+
+    if type(world_size) is list:
+        world_size = max(world_size)
+
+    if world_size != total_files:
+        raise ValueError(
+            f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. "
+            "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes."
+        )
+
+    # the groups are named differently in each stage
+    if zero_stage <= 2:
+        fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS
+    elif zero_stage == 3:
+        fp32_groups_key = FP32_FLAT_GROUPS
+    else:
+        raise ValueError(f"unknown zero stage {zero_stage}")
+
+    fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))]
+    return zero_stage, world_size, fp32_flat_groups
+
+
+def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters):
+    """
+    Returns fp32 state_dict reconstructed from ds checkpoint
+
+    Args:
+        - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are)
+
+    """
+    print(f"Processing zero checkpoint '{ds_checkpoint_dir}'")
+
+    optim_files = get_optim_files(ds_checkpoint_dir)
+    zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir)
+    print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}")
+
+    model_files = get_model_state_files(ds_checkpoint_dir)
+
+    zero_model_states = parse_model_states(model_files)
+    print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}')
+
+    if zero_stage <= 2:
+        return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                                          exclude_frozen_parameters)
+    elif zero_stage == 3:
+        return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                                          exclude_frozen_parameters)
+
+
+def _zero2_merge_frozen_params(state_dict, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+    frozen_param_fragments = zero_model_states[0].frozen_param_fragments
+
+    if debug:
+        num_elem = sum(s.numel() for s in frozen_param_shapes.values())
+        print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in frozen_param_fragments.values()])
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        state_dict[name] = frozen_param_fragments[name]
+
+        if debug:
+            print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _has_callable(obj, fn):
+    attr = getattr(obj, fn, None)
+    return callable(attr)
+
+
+def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+
+    # Reconstruction protocol:
+    #
+    # XXX: document this
+
+    if debug:
+        for i in range(world_size):
+            for j in range(len(fp32_flat_groups[0])):
+                print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}")
+
+    # XXX: memory usage doubles here (zero2)
+    num_param_groups = len(fp32_flat_groups[0])
+    merged_single_partition_of_fp32_groups = []
+    for i in range(num_param_groups):
+        merged_partitions = [sd[i] for sd in fp32_flat_groups]
+        full_single_fp32_vector = torch.cat(merged_partitions, 0)
+        merged_single_partition_of_fp32_groups.append(full_single_fp32_vector)
+    avail_numel = sum(
+        [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups])
+
+    if debug:
+        wanted_params = sum([len(shapes) for shapes in param_shapes])
+        wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes])
+        # not asserting if there is a mismatch due to possible padding
+        print(f"Have {avail_numel} numels to process.")
+        print(f"Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    total_numel = 0
+    total_params = 0
+    for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups):
+        offset = 0
+        avail_numel = full_single_fp32_vector.numel()
+        for name, shape in shapes.items():
+
+            unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape)
+            total_numel += unpartitioned_numel
+            total_params += 1
+
+            if debug:
+                print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+            state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape)
+            offset += unpartitioned_numel
+
+        # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and
+        # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex
+        # paddings performed in the code it's almost impossible to predict the exact numbers w/o the
+        # live optimizer object, so we are checking that the numbers are within the right range
+        align_to = 2 * world_size
+
+        def zero2_align(x):
+            return align_to * math.ceil(x / align_to)
+
+        if debug:
+            print(f"original offset={offset}, avail_numel={avail_numel}")
+
+        offset = zero2_align(offset)
+        avail_numel = zero2_align(avail_numel)
+
+        if debug:
+            print(f"aligned  offset={offset}, avail_numel={avail_numel}")
+
+        # Sanity check
+        if offset != avail_numel:
+            raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+    print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                               exclude_frozen_parameters):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    if not exclude_frozen_parameters:
+        _zero2_merge_frozen_params(state_dict, zero_model_states)
+
+    _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def zero3_partitioned_param_info(unpartitioned_numel, world_size):
+    remainder = unpartitioned_numel % world_size
+    padding_numel = (world_size - remainder) if remainder else 0
+    partitioned_numel = math.ceil(unpartitioned_numel / world_size)
+    return partitioned_numel, padding_numel
+
+
+def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    if debug:
+        for i in range(world_size):
+            num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values())
+            print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in zero_model_states[0].frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states)
+        state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape)
+
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+class GatheredTensor:
+    """
+    A pseudo tensor that collects partitioned weights.
+    It is more memory efficient when there are multiple groups.
+    """
+
+    def __init__(self, flat_groups, flat_groups_offset, offset, partitioned_numel, shape):
+        self.flat_groups = flat_groups
+        self.flat_groups_offset = flat_groups_offset
+        self.offset = offset
+        self.partitioned_numel = partitioned_numel
+        self.shape = shape
+        self.dtype = self.flat_groups[0][0].dtype
+
+    def contiguous(self):
+        """
+        Merge partitioned weights from flat_groups into a single tensor.
+        """
+        end_idx = self.offset + self.partitioned_numel
+        world_size = len(self.flat_groups)
+        pad_flat_param_chunks = []
+
+        for rank_i in range(world_size):
+            # for each rank, we need to collect weights from related group/groups
+            flat_groups_at_rank_i = self.flat_groups[rank_i]
+            start_group_id = None
+            end_group_id = None
+            for group_id in range(len(self.flat_groups_offset)):
+                if self.flat_groups_offset[group_id] <= self.offset < self.flat_groups_offset[group_id + 1]:
+                    start_group_id = group_id
+                if self.flat_groups_offset[group_id] < end_idx <= self.flat_groups_offset[group_id + 1]:
+                    end_group_id = group_id
+                    break
+            # collect weights from related group/groups
+            for group_id in range(start_group_id, end_group_id + 1):
+                flat_tensor = flat_groups_at_rank_i[group_id]
+                start_offset = self.offset - self.flat_groups_offset[group_id]
+                end_offset = min(end_idx, self.flat_groups_offset[group_id + 1]) - self.flat_groups_offset[group_id]
+                pad_flat_param_chunks.append(flat_tensor[start_offset:end_offset])
+
+        # collect weights from all ranks
+        pad_flat_param = torch.cat(pad_flat_param_chunks, dim=0)
+        param = pad_flat_param[:self.shape.numel()].view(self.shape).contiguous()
+        return param
+
+
+def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+    avail_numel = sum([flat_group.numel() for flat_group in fp32_flat_groups[0]]) * world_size
+
+    # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each
+    # param, re-consolidating each param, while dealing with padding if any
+
+    # merge list of dicts, preserving order
+    param_shapes = {k: v for d in param_shapes for k, v in d.items()}
+
+    if debug:
+        for i in range(world_size):
+            print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}")
+
+        wanted_params = len(param_shapes)
+        wanted_numel = sum(shape.numel() for shape in param_shapes.values())
+        # not asserting if there is a mismatch due to possible padding
+        avail_numel = fp32_flat_groups[0].numel() * world_size
+        print(f"Trainable params: Have {avail_numel} numels to process.")
+        print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    offset = 0
+    total_numel = 0
+    total_params = 0
+    flat_groups_offset = [0] + list(np.cumsum([flat_tensor.numel() for flat_tensor in fp32_flat_groups[0]]))
+    for name, shape in tqdm(param_shapes.items(), desc='Gathering sharded weights'):
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+        total_params += 1
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+        # memory efficient tensor
+        tensor = GatheredTensor(fp32_flat_groups, flat_groups_offset, offset, partitioned_numel, shape)
+        state_dict[name] = tensor
+        offset += partitioned_numel
+
+    offset *= world_size
+
+    # Sanity check
+    if offset != avail_numel:
+        raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+    print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                               exclude_frozen_parameters):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    if not exclude_frozen_parameters:
+        _zero3_merge_frozen_params(state_dict, world_size, zero_model_states)
+
+    _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def to_torch_tensor(state_dict, return_empty_tensor=False):
+    """
+    Convert state_dict of GatheredTensor to torch tensor
+    """
+    torch_state_dict = {}
+    converted_tensors = {}
+    for name, tensor in state_dict.items():
+        tensor_id = id(tensor)
+        if tensor_id in converted_tensors:  # shared tensors
+            shared_tensor = torch_state_dict[converted_tensors[tensor_id]]
+            torch_state_dict[name] = shared_tensor
+        else:
+            converted_tensors[tensor_id] = name
+            if return_empty_tensor:
+                torch_state_dict[name] = torch.empty(tensor.shape, dtype=tensor.dtype)
+            else:
+                torch_state_dict[name] = tensor.contiguous()
+    return torch_state_dict
+
+
+def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir,
+                                             tag=None,
+                                             exclude_frozen_parameters=False,
+                                             lazy_mode=False):
+    """
+    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with
+    ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example
+    via a model hub.
+
+    Args:
+        - ``checkpoint_dir``: path to the desired checkpoint folder
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14``
+        - ``exclude_frozen_parameters``: exclude frozen parameters
+        - ``lazy_mode``: get state_dict in lazy mode. It returns a dict of pesduo tensor instead of torch tensor, which is more memory efficient.
+          Convert the pesduo tensor to torch tensor by ``.contiguous()``
+
+    Returns:
+        - pytorch ``state_dict``
+
+    A typical usage might be ::
+
+        from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
+        # do the training and checkpoint saving
+        state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu
+        model = model.cpu() # move to cpu
+        model.load_state_dict(state_dict)
+        # submit to model hub or save the model to share with others
+
+    In this example the ``model`` will no longer be usable in the deepspeed context of the same
+    application. i.e. you will need to re-initialize the deepspeed engine, since
+    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+    If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead.
+
+    Note: the above usage may not work if your application doesn't have sufficient free CPU memory.
+    You may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with
+    the checkpoint. Or you can load state_dict in lazy mode ::
+
+        from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
+        state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, lazy_mode=True) # not on cpu
+        for name, lazy_tensor in state_dict.item():
+            tensor = lazy_tensor.contiguous()  # to cpu
+            print(name, tensor)
+            # del tensor to release memory if it no longer in use
+    """
+    if tag is None:
+        latest_path = os.path.join(checkpoint_dir, 'latest')
+        if os.path.isfile(latest_path):
+            with open(latest_path, 'r') as fd:
+                tag = fd.read().strip()
+        else:
+            raise ValueError(f"Unable to find 'latest' file at {latest_path}")
+
+    ds_checkpoint_dir = os.path.join(checkpoint_dir, tag)
+
+    if not os.path.isdir(ds_checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist")
+
+    state_dict = _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters)
+    if lazy_mode:
+        return state_dict
+    else:
+        return to_torch_tensor(state_dict)
+
+
+def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir,
+                                               output_dir,
+                                               max_shard_size="5GB",
+                                               safe_serialization=False,
+                                               tag=None,
+                                               exclude_frozen_parameters=False):
+    """
+    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be
+    loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed.
+
+    Args:
+        - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+        - ``output_dir``: directory to the pytorch fp32 state_dict output files
+        - ``max_shard_size``: the maximum size for a checkpoint before being sharded, default value is 5GB
+        - ``safe_serialization``:  whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+        - ``exclude_frozen_parameters``: exclude frozen parameters
+    """
+
+    # Dependency pre-check
+    if safe_serialization:
+        try:
+            from safetensors.torch import save_file
+        except ImportError:
+            print('If you want to use `safe_serialization`, please `pip install safetensors`')
+            raise
+    if max_shard_size is not None:
+        try:
+            from huggingface_hub import split_torch_state_dict_into_shards
+        except ImportError:
+            print('If you want to use `max_shard_size`, please `pip install huggingface_hub`')
+            raise
+
+    # Convert zero checkpoint to state_dict
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir,
+                                                          tag,
+                                                          exclude_frozen_parameters,
+                                                          lazy_mode=True)
+
+    # Shard the model if it is too big.
+    weights_name = "model.safetensors" if safe_serialization else "pytorch_model.bin"
+    if max_shard_size is not None:
+        filename_pattern = weights_name.replace(".bin", "{suffix}.bin").replace(".safetensors", "{suffix}.safetensors")
+        # an memory-efficient approach for sharding
+        empty_state_dict = to_torch_tensor(state_dict, return_empty_tensor=True)
+        state_dict_split = split_torch_state_dict_into_shards(empty_state_dict,
+                                                              filename_pattern=filename_pattern,
+                                                              max_shard_size=max_shard_size)
+    else:
+        from collections import namedtuple
+        StateDictSplit = namedtuple("StateDictSplit", ["is_sharded", "filename_to_tensors"])
+        state_dict_split = StateDictSplit(is_sharded=False,
+                                          filename_to_tensors={weights_name: list(state_dict.keys())})
+
+    # Save the model by shard
+    os.makedirs(output_dir, exist_ok=True)
+    filename_to_tensors = state_dict_split.filename_to_tensors.items()
+    for shard_file, tensors in tqdm(filename_to_tensors, desc="Saving checkpoint shards"):
+        shard_state_dict = {tensor_name: state_dict[tensor_name] for tensor_name in tensors}
+        shard_state_dict = to_torch_tensor(shard_state_dict)
+        output_path = os.path.join(output_dir, shard_file)
+        if safe_serialization:
+            save_file(shard_state_dict, output_path, metadata={"format": "pt"})
+        else:
+            torch.save(shard_state_dict, output_path)
+        # release the memory of current shard
+        for tensor_name in list(shard_state_dict.keys()):
+            del state_dict[tensor_name]
+            del shard_state_dict[tensor_name]
+        del shard_state_dict
+        gc.collect()
+
+    # Save index if sharded
+    if state_dict_split.is_sharded:
+        index = {
+            "metadata": state_dict_split.metadata,
+            "weight_map": state_dict_split.tensor_to_filename,
+        }
+        save_index_file = "model.safetensors.index.json" if safe_serialization else "pytorch_model.bin.index.json"
+        save_index_file = os.path.join(output_dir, save_index_file)
+        with open(save_index_file, "w", encoding="utf-8") as f:
+            content = json.dumps(index, indent=2, sort_keys=True) + "\n"
+            f.write(content)
+
+
+def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None):
+    """
+    1. Put the provided model to cpu
+    2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict``
+    3. Load it into the provided model
+
+    Args:
+        - ``model``: the model object to update
+        - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+
+    Returns:
+        - ``model`: modified model
+
+    Make sure you have plenty of CPU memory available before you call this function. If you don't
+    have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it
+    conveniently placed for you in the checkpoint folder.
+
+    A typical usage might be ::
+
+        from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint
+        model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir)
+        # submit to model hub or save the model to share with others
+
+    Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context
+    of the same application. i.e. you will need to re-initialize the deepspeed engine, since
+    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+    """
+    logger.info(f"Extracting fp32 weights")
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag)
+
+    logger.info(f"Overwriting model with fp32 weights")
+    model = model.cpu()
+    model.load_state_dict(state_dict, strict=False)
+
+    return model
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser()
+    parser.add_argument("checkpoint_dir",
+                        type=str,
+                        help="path to the desired checkpoint folder, e.g., path/checkpoint-12")
+    parser.add_argument("output_dir",
+                        type=str,
+                        help="directory to the pytorch fp32 state_dict output files"
+                        "(e.g. path/checkpoint-12-output/)")
+    parser.add_argument(
+        "--max_shard_size",
+        type=str,
+        default="5GB",
+        help="The maximum size for a checkpoint before being sharded. Checkpoints shard will then be each of size"
+        "lower than this size. If expressed as a string, needs to be digits followed by a unit (like `5MB`"
+        "We default it to 5GB in order for models to be able to run easily on free-tier google colab instances"
+        "without CPU OOM issues.")
+    parser.add_argument(
+        "--safe_serialization",
+        default=False,
+        action='store_true',
+        help="Whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).")
+    parser.add_argument("-t",
+                        "--tag",
+                        type=str,
+                        default=None,
+                        help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1")
+    parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters")
+    parser.add_argument("-d", "--debug", action='store_true', help="enable debug")
+    args = parser.parse_args()
+
+    debug = args.debug
+
+    convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir,
+                                               args.output_dir,
+                                               max_shard_size=args.max_shard_size,
+                                               safe_serialization=args.safe_serialization,
+                                               tag=args.tag,
+                                               exclude_frozen_parameters=args.exclude_frozen_parameters)

3b-w-cot+/checkpoint-488/added_tokens.json

@@ -0,0 +1,24 @@
+{
+  "</tool_call>": 151658,
+  "<tool_call>": 151657,
+  "<|box_end|>": 151649,
+  "<|box_start|>": 151648,
+  "<|endoftext|>": 151643,
+  "<|file_sep|>": 151664,
+  "<|fim_middle|>": 151660,
+  "<|fim_pad|>": 151662,
+  "<|fim_prefix|>": 151659,
+  "<|fim_suffix|>": 151661,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644,
+  "<|image_pad|>": 151655,
+  "<|object_ref_end|>": 151647,
+  "<|object_ref_start|>": 151646,
+  "<|quad_end|>": 151651,
+  "<|quad_start|>": 151650,
+  "<|repo_name|>": 151663,
+  "<|video_pad|>": 151656,
+  "<|vision_end|>": 151653,
+  "<|vision_pad|>": 151654,
+  "<|vision_start|>": 151652
+}

3b-w-cot+/checkpoint-488/config.json

@@ -0,0 +1,28 @@
+{
+  "_name_or_path": "ckpt/3b-w-cot/checkpoint-747",
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "eos_token_id": 151645,
+  "hidden_act": "silu",
+  "hidden_size": 2048,
+  "initializer_range": 0.02,
+  "intermediate_size": 11008,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 70,
+  "model_type": "qwen2",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 36,
+  "num_key_value_heads": 2,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.49.0",
+  "use_cache": false,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}

3b-w-cot+/checkpoint-488/generation_config.json

@@ -0,0 +1,14 @@
+{
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "repetition_penalty": 1.05,
+  "temperature": 0.7,
+  "top_k": 20,
+  "top_p": 0.8,
+  "transformers_version": "4.49.0"
+}

3b-w-cot+/checkpoint-488/global_step487/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:42d46bdbf299711fbd8125eb8acb726e12a4f88d9fcccb46d9547f789bc6ba13
+size 9306058322

3b-w-cot+/checkpoint-488/global_step487/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e1187b7d477a60d412e23458de4b2cd75b8c1b327e898b5ad66e98b037b7be12
+size 9306060690

3b-w-cot+/checkpoint-488/global_step487/mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6483efeebfee3a33c5ad10ecf28e2cf27546b77b1c2be25e152137ec67effc6f
+size 6171993592

3b-w-cot+/checkpoint-488/latest

@@ -0,0 +1 @@
+global_step487

3b-w-cot+/checkpoint-488/model-00001-of-00002.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fe067b847f63d4cf5dc966e0b39c3c0ae6b5c580b4fc58449eeca26fa528b266
+size 4957560304

3b-w-cot+/checkpoint-488/model-00002-of-00002.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b2297c2359849d5acc982147817db1c90cf1170f8f42c305f65c42e9a49bf56e
+size 1836696752

3b-w-cot+/checkpoint-488/model.safetensors.index.json

@@ -0,0 +1,442 @@
+{
+  "metadata": {
+    "total_size": 6794207232
+  },
+  "weight_map": {
+    "lm_head.weight": "model-00002-of-00002.safetensors",
+    "model.embed_tokens.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.28.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.28.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.28.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.28.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.28.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.28.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.28.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.29.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.30.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.norm.weight": "model-00002-of-00002.safetensors"
+  }
+}

3b-w-cot+/checkpoint-488/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3dcb161b22b2558dbf7e3f8c871050cec383d11a40423fab11f18d5e630639bf
+size 14512

3b-w-cot+/checkpoint-488/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d50af6aef769414a6f28fa1b1bc51ce707dc8ecd15474e03f99a2f10fde086be
+size 14512

3b-w-cot+/checkpoint-488/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6d8b2a59c30f5e09b1d7ce944fea889fdfc7000e147a68a8ad08ea9263213eb2
+size 1064

3b-w-cot+/checkpoint-488/special_tokens_map.json

@@ -0,0 +1,31 @@
+{
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "eos_token": {
+    "content": "<|im_end|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

3b-w-cot+/checkpoint-488/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9c5ae00e602b8860cbd784ba82a8aa14e8feecec692e7076590d014d7b7fdafa
+size 11421896

3b-w-cot+/checkpoint-488/tokenizer_config.json

@@ -0,0 +1,208 @@
+{
+  "add_bos_token": false,
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151645": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151646": {
+      "content": "<|object_ref_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151647": {
+      "content": "<|object_ref_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151648": {
+      "content": "<|box_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151649": {
+      "content": "<|box_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151650": {
+      "content": "<|quad_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151651": {
+      "content": "<|quad_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151652": {
+      "content": "<|vision_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151653": {
+      "content": "<|vision_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151654": {
+      "content": "<|vision_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151655": {
+      "content": "<|image_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151656": {
+      "content": "<|video_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151657": {
+      "content": "<tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151658": {
+      "content": "</tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151659": {
+      "content": "<|fim_prefix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151660": {
+      "content": "<|fim_middle|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151661": {
+      "content": "<|fim_suffix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151662": {
+      "content": "<|fim_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151663": {
+      "content": "<|repo_name|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151664": {
+      "content": "<|file_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    }
+  },
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "bos_token": null,
+  "chat_template": "{%- if tools %}\n    {{- '<|im_start|>system\\n' }}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- messages[0]['content'] }}\n    {%- else %}\n        {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}\n    {%- endif %}\n    {{- \"\\n\\n# Tools\\n\\nYou may call one or more functions to assist with the user query.\\n\\nYou are provided with function signatures within <tools></tools> XML tags:\\n<tools>\" }}\n    {%- for tool in tools %}\n        {{- \"\\n\" }}\n        {{- tool | tojson }}\n    {%- endfor %}\n    {{- \"\\n</tools>\\n\\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\\n<tool_call>\\n{\\\"name\\\": <function-name>, \\\"arguments\\\": <args-json-object>}\\n</tool_call><|im_end|>\\n\" }}\n{%- else %}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- '<|im_start|>system\\n' + messages[0]['content'] + '<|im_end|>\\n' }}\n    {%- else %}\n        {{- '<|im_start|>system\\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\\n' }}\n    {%- endif %}\n{%- endif %}\n{%- for message in messages %}\n    {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) or (message.role == \"assistant\" and not message.tool_calls) %}\n        {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {{- '<|im_start|>' + message.role }}\n        {%- if message.content %}\n            {{- '\\n' + message.content }}\n        {%- endif %}\n        {%- for tool_call in message.tool_calls %}\n            {%- if tool_call.function is defined %}\n                {%- set tool_call = tool_call.function %}\n            {%- endif %}\n            {{- '\\n<tool_call>\\n{\"name\": \"' }}\n            {{- tool_call.name }}\n            {{- '\", \"arguments\": ' }}\n            {{- tool_call.arguments | tojson }}\n            {{- '}\\n</tool_call>' }}\n        {%- endfor %}\n        {{- '<|im_end|>\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != \"tool\") %}\n            {{- '<|im_start|>user' }}\n        {%- endif %}\n        {{- '\\n<tool_response>\\n' }}\n        {{- message.content }}\n        {{- '\\n</tool_response>' }}\n        {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n            {{- '<|im_end|>\\n' }}\n        {%- endif %}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {{- '<|im_start|>assistant\\n' }}\n{%- endif %}\n",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": {},
+  "model_max_length": 131072,
+  "pad_token": "<|endoftext|>",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

3b-w-cot+/checkpoint-488/trainer_state.json

@@ -0,0 +1,3497 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.9948822927328558,
+  "eval_steps": 82,
+  "global_step": 488,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0040941658137154556,
+      "grad_norm": 8.007163047790527,
+      "learning_rate": 6.666666666666667e-07,
+      "loss": 0.477,
+      "step": 1
+    },
+    {
+      "epoch": 0.0040941658137154556,
+      "eval_loss": 0.8367487192153931,
+      "eval_runtime": 4.4844,
+      "eval_samples_per_second": 17.617,
+      "eval_steps_per_second": 2.23,
+      "step": 1
+    },
+    {
+      "epoch": 0.008188331627430911,
+      "grad_norm": 8.800883293151855,
+      "learning_rate": 1.3333333333333334e-06,
+      "loss": 0.6282,
+      "step": 2
+    },
+    {
+      "epoch": 0.012282497441146366,
+      "grad_norm": 9.306445121765137,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.6202,
+      "step": 3
+    },
+    {
+      "epoch": 0.016376663254861822,
+      "grad_norm": 10.017245292663574,
+      "learning_rate": 2.666666666666667e-06,
+      "loss": 0.6161,
+      "step": 4
+    },
+    {
+      "epoch": 0.02047082906857728,
+      "grad_norm": 9.14148235321045,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.6081,
+      "step": 5
+    },
+    {
+      "epoch": 0.02456499488229273,
+      "grad_norm": 8.813340187072754,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.5697,
+      "step": 6
+    },
+    {
+      "epoch": 0.028659160696008188,
+      "grad_norm": 6.7533745765686035,
+      "learning_rate": 4.666666666666667e-06,
+      "loss": 0.4834,
+      "step": 7
+    },
+    {
+      "epoch": 0.032753326509723645,
+      "grad_norm": 6.1987481117248535,
+      "learning_rate": 5.333333333333334e-06,
+      "loss": 0.4857,
+      "step": 8
+    },
+    {
+      "epoch": 0.0368474923234391,
+      "grad_norm": 2.4827005863189697,
+      "learning_rate": 6e-06,
+      "loss": 0.3713,
+      "step": 9
+    },
+    {
+      "epoch": 0.04094165813715456,
+      "grad_norm": 2.163064956665039,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.3284,
+      "step": 10
+    },
+    {
+      "epoch": 0.04503582395087001,
+      "grad_norm": 1.9997942447662354,
+      "learning_rate": 7.333333333333333e-06,
+      "loss": 0.3289,
+      "step": 11
+    },
+    {
+      "epoch": 0.04912998976458546,
+      "grad_norm": 2.7956204414367676,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.3208,
+      "step": 12
+    },
+    {
+      "epoch": 0.05322415557830092,
+      "grad_norm": 2.8886733055114746,
+      "learning_rate": 8.666666666666668e-06,
+      "loss": 0.3123,
+      "step": 13
+    },
+    {
+      "epoch": 0.057318321392016376,
+      "grad_norm": 2.217071771621704,
+      "learning_rate": 9.333333333333334e-06,
+      "loss": 0.2881,
+      "step": 14
+    },
+    {
+      "epoch": 0.06141248720573183,
+      "grad_norm": 1.9985229969024658,
+      "learning_rate": 1e-05,
+      "loss": 0.283,
+      "step": 15
+    },
+    {
+      "epoch": 0.06550665301944729,
+      "grad_norm": 1.8881174325942993,
+      "learning_rate": 1.0666666666666667e-05,
+      "loss": 0.2616,
+      "step": 16
+    },
+    {
+      "epoch": 0.06960081883316274,
+      "grad_norm": 1.9551236629486084,
+      "learning_rate": 1.1333333333333334e-05,
+      "loss": 0.2894,
+      "step": 17
+    },
+    {
+      "epoch": 0.0736949846468782,
+      "grad_norm": 1.8677968978881836,
+      "learning_rate": 1.2e-05,
+      "loss": 0.2328,
+      "step": 18
+    },
+    {
+      "epoch": 0.07778915046059365,
+      "grad_norm": 1.9170935153961182,
+      "learning_rate": 1.2666666666666667e-05,
+      "loss": 0.2577,
+      "step": 19
+    },
+    {
+      "epoch": 0.08188331627430911,
+      "grad_norm": 2.189279794692993,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.2555,
+      "step": 20
+    },
+    {
+      "epoch": 0.08597748208802457,
+      "grad_norm": 3.202075242996216,
+      "learning_rate": 1.4e-05,
+      "loss": 0.2647,
+      "step": 21
+    },
+    {
+      "epoch": 0.09007164790174002,
+      "grad_norm": 2.774186372756958,
+      "learning_rate": 1.4666666666666666e-05,
+      "loss": 0.2531,
+      "step": 22
+    },
+    {
+      "epoch": 0.09416581371545547,
+      "grad_norm": 2.0601961612701416,
+      "learning_rate": 1.5333333333333334e-05,
+      "loss": 0.2654,
+      "step": 23
+    },
+    {
+      "epoch": 0.09825997952917093,
+      "grad_norm": 1.781900405883789,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.2358,
+      "step": 24
+    },
+    {
+      "epoch": 0.1023541453428864,
+      "grad_norm": 1.8549216985702515,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.2362,
+      "step": 25
+    },
+    {
+      "epoch": 0.10644831115660185,
+      "grad_norm": 2.1376802921295166,
+      "learning_rate": 1.7333333333333336e-05,
+      "loss": 0.2334,
+      "step": 26
+    },
+    {
+      "epoch": 0.1105424769703173,
+      "grad_norm": 2.134582281112671,
+      "learning_rate": 1.8e-05,
+      "loss": 0.267,
+      "step": 27
+    },
+    {
+      "epoch": 0.11463664278403275,
+      "grad_norm": 1.6425909996032715,
+      "learning_rate": 1.866666666666667e-05,
+      "loss": 0.2184,
+      "step": 28
+    },
+    {
+      "epoch": 0.1187308085977482,
+      "grad_norm": 1.601938009262085,
+      "learning_rate": 1.9333333333333333e-05,
+      "loss": 0.256,
+      "step": 29
+    },
+    {
+      "epoch": 0.12282497441146366,
+      "grad_norm": 1.626160740852356,
+      "learning_rate": 2e-05,
+      "loss": 0.2548,
+      "step": 30
+    },
+    {
+      "epoch": 0.1269191402251791,
+      "grad_norm": 1.6894042491912842,
+      "learning_rate": 1.999989986294826e-05,
+      "loss": 0.2438,
+      "step": 31
+    },
+    {
+      "epoch": 0.13101330603889458,
+      "grad_norm": 1.6300302743911743,
+      "learning_rate": 1.9999599453798523e-05,
+      "loss": 0.2273,
+      "step": 32
+    },
+    {
+      "epoch": 0.13510747185261002,
+      "grad_norm": 1.6301401853561401,
+      "learning_rate": 1.999909877856721e-05,
+      "loss": 0.2539,
+      "step": 33
+    },
+    {
+      "epoch": 0.13920163766632548,
+      "grad_norm": 1.6974273920059204,
+      "learning_rate": 1.9998397847281548e-05,
+      "loss": 0.256,
+      "step": 34
+    },
+    {
+      "epoch": 0.14329580348004095,
+      "grad_norm": 1.5356749296188354,
+      "learning_rate": 1.9997496673979375e-05,
+      "loss": 0.2278,
+      "step": 35
+    },
+    {
+      "epoch": 0.1473899692937564,
+      "grad_norm": 1.6304699182510376,
+      "learning_rate": 1.9996395276708856e-05,
+      "loss": 0.2488,
+      "step": 36
+    },
+    {
+      "epoch": 0.15148413510747186,
+      "grad_norm": 1.6241912841796875,
+      "learning_rate": 1.999509367752813e-05,
+      "loss": 0.2407,
+      "step": 37
+    },
+    {
+      "epoch": 0.1555783009211873,
+      "grad_norm": 1.6954501867294312,
+      "learning_rate": 1.9993591902504854e-05,
+      "loss": 0.2279,
+      "step": 38
+    },
+    {
+      "epoch": 0.15967246673490276,
+      "grad_norm": 1.5775200128555298,
+      "learning_rate": 1.9991889981715696e-05,
+      "loss": 0.2443,
+      "step": 39
+    },
+    {
+      "epoch": 0.16376663254861823,
+      "grad_norm": 1.6417704820632935,
+      "learning_rate": 1.9989987949245725e-05,
+      "loss": 0.2498,
+      "step": 40
+    },
+    {
+      "epoch": 0.16786079836233367,
+      "grad_norm": 1.6866360902786255,
+      "learning_rate": 1.9987885843187717e-05,
+      "loss": 0.2496,
+      "step": 41
+    },
+    {
+      "epoch": 0.17195496417604914,
+      "grad_norm": 1.6979321241378784,
+      "learning_rate": 1.9985583705641418e-05,
+      "loss": 0.2721,
+      "step": 42
+    },
+    {
+      "epoch": 0.17604912998976457,
+      "grad_norm": 1.676047682762146,
+      "learning_rate": 1.9983081582712684e-05,
+      "loss": 0.2506,
+      "step": 43
+    },
+    {
+      "epoch": 0.18014329580348004,
+      "grad_norm": 1.7262601852416992,
+      "learning_rate": 1.998037952451255e-05,
+      "loss": 0.2371,
+      "step": 44
+    },
+    {
+      "epoch": 0.1842374616171955,
+      "grad_norm": 1.4194152355194092,
+      "learning_rate": 1.9977477585156252e-05,
+      "loss": 0.2619,
+      "step": 45
+    },
+    {
+      "epoch": 0.18833162743091095,
+      "grad_norm": 1.5654889345169067,
+      "learning_rate": 1.9974375822762117e-05,
+      "loss": 0.2299,
+      "step": 46
+    },
+    {
+      "epoch": 0.19242579324462641,
+      "grad_norm": 1.844489336013794,
+      "learning_rate": 1.9971074299450414e-05,
+      "loss": 0.2692,
+      "step": 47
+    },
+    {
+      "epoch": 0.19651995905834185,
+      "grad_norm": 1.5128370523452759,
+      "learning_rate": 1.9967573081342103e-05,
+      "loss": 0.2589,
+      "step": 48
+    },
+    {
+      "epoch": 0.20061412487205732,
+      "grad_norm": 1.5005507469177246,
+      "learning_rate": 1.9963872238557516e-05,
+      "loss": 0.2578,
+      "step": 49
+    },
+    {
+      "epoch": 0.2047082906857728,
+      "grad_norm": 1.5974067449569702,
+      "learning_rate": 1.9959971845214953e-05,
+      "loss": 0.2494,
+      "step": 50
+    },
+    {
+      "epoch": 0.20880245649948823,
+      "grad_norm": 1.5728641748428345,
+      "learning_rate": 1.9955871979429188e-05,
+      "loss": 0.2496,
+      "step": 51
+    },
+    {
+      "epoch": 0.2128966223132037,
+      "grad_norm": 1.5953929424285889,
+      "learning_rate": 1.9951572723309918e-05,
+      "loss": 0.2429,
+      "step": 52
+    },
+    {
+      "epoch": 0.21699078812691913,
+      "grad_norm": 1.7769081592559814,
+      "learning_rate": 1.9947074162960113e-05,
+      "loss": 0.252,
+      "step": 53
+    },
+    {
+      "epoch": 0.2210849539406346,
+      "grad_norm": 1.6964116096496582,
+      "learning_rate": 1.9942376388474282e-05,
+      "loss": 0.2651,
+      "step": 54
+    },
+    {
+      "epoch": 0.22517911975435004,
+      "grad_norm": 1.5599926710128784,
+      "learning_rate": 1.993747949393668e-05,
+      "loss": 0.2193,
+      "step": 55
+    },
+    {
+      "epoch": 0.2292732855680655,
+      "grad_norm": 1.524835467338562,
+      "learning_rate": 1.9932383577419432e-05,
+      "loss": 0.2361,
+      "step": 56
+    },
+    {
+      "epoch": 0.23336745138178097,
+      "grad_norm": 1.6240477561950684,
+      "learning_rate": 1.992708874098054e-05,
+      "loss": 0.2563,
+      "step": 57
+    },
+    {
+      "epoch": 0.2374616171954964,
+      "grad_norm": 1.6357301473617554,
+      "learning_rate": 1.9921595090661872e-05,
+      "loss": 0.2456,
+      "step": 58
+    },
+    {
+      "epoch": 0.24155578300921188,
+      "grad_norm": 1.5650979280471802,
+      "learning_rate": 1.991590273648702e-05,
+      "loss": 0.2512,
+      "step": 59
+    },
+    {
+      "epoch": 0.24564994882292732,
+      "grad_norm": 1.4107614755630493,
+      "learning_rate": 1.9910011792459086e-05,
+      "loss": 0.2539,
+      "step": 60
+    },
+    {
+      "epoch": 0.24974411463664278,
+      "grad_norm": 1.3979556560516357,
+      "learning_rate": 1.9903922376558432e-05,
+      "loss": 0.2348,
+      "step": 61
+    },
+    {
+      "epoch": 0.2538382804503582,
+      "grad_norm": 1.4066411256790161,
+      "learning_rate": 1.989763461074029e-05,
+      "loss": 0.2419,
+      "step": 62
+    },
+    {
+      "epoch": 0.2579324462640737,
+      "grad_norm": 1.533858060836792,
+      "learning_rate": 1.989114862093232e-05,
+      "loss": 0.252,
+      "step": 63
+    },
+    {
+      "epoch": 0.26202661207778916,
+      "grad_norm": 1.4007140398025513,
+      "learning_rate": 1.9884464537032103e-05,
+      "loss": 0.2379,
+      "step": 64
+    },
+    {
+      "epoch": 0.2661207778915046,
+      "grad_norm": 1.3841203451156616,
+      "learning_rate": 1.9877582492904533e-05,
+      "loss": 0.2377,
+      "step": 65
+    },
+    {
+      "epoch": 0.27021494370522003,
+      "grad_norm": 1.274598479270935,
+      "learning_rate": 1.9870502626379127e-05,
+      "loss": 0.2364,
+      "step": 66
+    },
+    {
+      "epoch": 0.2743091095189355,
+      "grad_norm": 1.59529447555542,
+      "learning_rate": 1.9863225079247286e-05,
+      "loss": 0.2647,
+      "step": 67
+    },
+    {
+      "epoch": 0.27840327533265097,
+      "grad_norm": 1.4986138343811035,
+      "learning_rate": 1.985574999725943e-05,
+      "loss": 0.264,
+      "step": 68
+    },
+    {
+      "epoch": 0.28249744114636643,
+      "grad_norm": 1.4692142009735107,
+      "learning_rate": 1.9848077530122083e-05,
+      "loss": 0.2475,
+      "step": 69
+    },
+    {
+      "epoch": 0.2865916069600819,
+      "grad_norm": 1.5602682828903198,
+      "learning_rate": 1.9840207831494903e-05,
+      "loss": 0.244,
+      "step": 70
+    },
+    {
+      "epoch": 0.2906857727737973,
+      "grad_norm": 1.5084633827209473,
+      "learning_rate": 1.983214105898757e-05,
+      "loss": 0.2442,
+      "step": 71
+    },
+    {
+      "epoch": 0.2947799385875128,
+      "grad_norm": 1.451074242591858,
+      "learning_rate": 1.9823877374156647e-05,
+      "loss": 0.2673,
+      "step": 72
+    },
+    {
+      "epoch": 0.29887410440122825,
+      "grad_norm": 1.4784746170043945,
+      "learning_rate": 1.9815416942502346e-05,
+      "loss": 0.2441,
+      "step": 73
+    },
+    {
+      "epoch": 0.3029682702149437,
+      "grad_norm": 1.368421196937561,
+      "learning_rate": 1.98067599334652e-05,
+      "loss": 0.2439,
+      "step": 74
+    },
+    {
+      "epoch": 0.3070624360286592,
+      "grad_norm": 1.4379913806915283,
+      "learning_rate": 1.979790652042268e-05,
+      "loss": 0.2238,
+      "step": 75
+    },
+    {
+      "epoch": 0.3111566018423746,
+      "grad_norm": 1.5823594331741333,
+      "learning_rate": 1.978885688068572e-05,
+      "loss": 0.2725,
+      "step": 76
+    },
+    {
+      "epoch": 0.31525076765609006,
+      "grad_norm": 1.5007753372192383,
+      "learning_rate": 1.9779611195495177e-05,
+      "loss": 0.2375,
+      "step": 77
+    },
+    {
+      "epoch": 0.3193449334698055,
+      "grad_norm": 1.473763346672058,
+      "learning_rate": 1.977016965001817e-05,
+      "loss": 0.2611,
+      "step": 78
+    },
+    {
+      "epoch": 0.323439099283521,
+      "grad_norm": 1.7358968257904053,
+      "learning_rate": 1.976053243334442e-05,
+      "loss": 0.2623,
+      "step": 79
+    },
+    {
+      "epoch": 0.32753326509723646,
+      "grad_norm": 1.5595204830169678,
+      "learning_rate": 1.9750699738482403e-05,
+      "loss": 0.2676,
+      "step": 80
+    },
+    {
+      "epoch": 0.33162743091095187,
+      "grad_norm": 1.685775637626648,
+      "learning_rate": 1.9740671762355548e-05,
+      "loss": 0.267,
+      "step": 81
+    },
+    {
+      "epoch": 0.33572159672466734,
+      "grad_norm": 1.3360930681228638,
+      "learning_rate": 1.973044870579824e-05,
+      "loss": 0.2339,
+      "step": 82
+    },
+    {
+      "epoch": 0.33572159672466734,
+      "eval_loss": 0.27538755536079407,
+      "eval_runtime": 5.5418,
+      "eval_samples_per_second": 14.255,
+      "eval_steps_per_second": 1.804,
+      "step": 82
+    },
+    {
+      "epoch": 0.3398157625383828,
+      "grad_norm": 1.7223079204559326,
+      "learning_rate": 1.972003077355183e-05,
+      "loss": 0.2843,
+      "step": 83
+    },
+    {
+      "epoch": 0.34390992835209827,
+      "grad_norm": 1.5389612913131714,
+      "learning_rate": 1.9709418174260523e-05,
+      "loss": 0.2604,
+      "step": 84
+    },
+    {
+      "epoch": 0.34800409416581374,
+      "grad_norm": 1.3990126848220825,
+      "learning_rate": 1.9698611120467196e-05,
+      "loss": 0.2588,
+      "step": 85
+    },
+    {
+      "epoch": 0.35209825997952915,
+      "grad_norm": 1.3090544939041138,
+      "learning_rate": 1.9687609828609156e-05,
+      "loss": 0.2264,
+      "step": 86
+    },
+    {
+      "epoch": 0.3561924257932446,
+      "grad_norm": 1.3547587394714355,
+      "learning_rate": 1.9676414519013782e-05,
+      "loss": 0.2436,
+      "step": 87
+    },
+    {
+      "epoch": 0.3602865916069601,
+      "grad_norm": 1.7848924398422241,
+      "learning_rate": 1.966502541589414e-05,
+      "loss": 0.2617,
+      "step": 88
+    },
+    {
+      "epoch": 0.36438075742067555,
+      "grad_norm": 1.450150728225708,
+      "learning_rate": 1.965344274734447e-05,
+      "loss": 0.2456,
+      "step": 89
+    },
+    {
+      "epoch": 0.368474923234391,
+      "grad_norm": 1.5955520868301392,
+      "learning_rate": 1.9641666745335626e-05,
+      "loss": 0.2571,
+      "step": 90
+    },
+    {
+      "epoch": 0.3725690890481064,
+      "grad_norm": 1.4538159370422363,
+      "learning_rate": 1.9629697645710432e-05,
+      "loss": 0.2552,
+      "step": 91
+    },
+    {
+      "epoch": 0.3766632548618219,
+      "grad_norm": 1.4258908033370972,
+      "learning_rate": 1.961753568817896e-05,
+      "loss": 0.2402,
+      "step": 92
+    },
+    {
+      "epoch": 0.38075742067553736,
+      "grad_norm": 1.463593602180481,
+      "learning_rate": 1.9605181116313725e-05,
+      "loss": 0.2612,
+      "step": 93
+    },
+    {
+      "epoch": 0.38485158648925283,
+      "grad_norm": 1.4867630004882812,
+      "learning_rate": 1.9592634177544803e-05,
+      "loss": 0.2272,
+      "step": 94
+    },
+    {
+      "epoch": 0.3889457523029683,
+      "grad_norm": 1.4335721731185913,
+      "learning_rate": 1.957989512315489e-05,
+      "loss": 0.2562,
+      "step": 95
+    },
+    {
+      "epoch": 0.3930399181166837,
+      "grad_norm": 1.3976503610610962,
+      "learning_rate": 1.9566964208274254e-05,
+      "loss": 0.2527,
+      "step": 96
+    },
+    {
+      "epoch": 0.3971340839303992,
+      "grad_norm": 1.455560564994812,
+      "learning_rate": 1.9553841691875632e-05,
+      "loss": 0.2617,
+      "step": 97
+    },
+    {
+      "epoch": 0.40122824974411464,
+      "grad_norm": 1.319207787513733,
+      "learning_rate": 1.9540527836769047e-05,
+      "loss": 0.251,
+      "step": 98
+    },
+    {
+      "epoch": 0.4053224155578301,
+      "grad_norm": 1.3503410816192627,
+      "learning_rate": 1.9527022909596537e-05,
+      "loss": 0.253,
+      "step": 99
+    },
+    {
+      "epoch": 0.4094165813715456,
+      "grad_norm": 1.472091555595398,
+      "learning_rate": 1.951332718082682e-05,
+      "loss": 0.2935,
+      "step": 100
+    },
+    {
+      "epoch": 0.413510747185261,
+      "grad_norm": 1.4811311960220337,
+      "learning_rate": 1.9499440924749878e-05,
+      "loss": 0.261,
+      "step": 101
+    },
+    {
+      "epoch": 0.41760491299897645,
+      "grad_norm": 1.595140814781189,
+      "learning_rate": 1.9485364419471454e-05,
+      "loss": 0.2703,
+      "step": 102
+    },
+    {
+      "epoch": 0.4216990788126919,
+      "grad_norm": 1.3853557109832764,
+      "learning_rate": 1.9471097946907506e-05,
+      "loss": 0.2345,
+      "step": 103
+    },
+    {
+      "epoch": 0.4257932446264074,
+      "grad_norm": 1.5886335372924805,
+      "learning_rate": 1.9456641792778527e-05,
+      "loss": 0.269,
+      "step": 104
+    },
+    {
+      "epoch": 0.42988741044012285,
+      "grad_norm": 1.3403784036636353,
+      "learning_rate": 1.9441996246603848e-05,
+      "loss": 0.234,
+      "step": 105
+    },
+    {
+      "epoch": 0.43398157625383826,
+      "grad_norm": 1.4540488719940186,
+      "learning_rate": 1.9427161601695833e-05,
+      "loss": 0.246,
+      "step": 106
+    },
+    {
+      "epoch": 0.43807574206755373,
+      "grad_norm": 1.5493131875991821,
+      "learning_rate": 1.9412138155154e-05,
+      "loss": 0.2284,
+      "step": 107
+    },
+    {
+      "epoch": 0.4421699078812692,
+      "grad_norm": 1.290971279144287,
+      "learning_rate": 1.9396926207859085e-05,
+      "loss": 0.2392,
+      "step": 108
+    },
+    {
+      "epoch": 0.44626407369498466,
+      "grad_norm": 1.4140104055404663,
+      "learning_rate": 1.9381526064466995e-05,
+      "loss": 0.2554,
+      "step": 109
+    },
+    {
+      "epoch": 0.4503582395087001,
+      "grad_norm": 1.7200373411178589,
+      "learning_rate": 1.9365938033402715e-05,
+      "loss": 0.2359,
+      "step": 110
+    },
+    {
+      "epoch": 0.45445240532241554,
+      "grad_norm": 1.2797805070877075,
+      "learning_rate": 1.9350162426854152e-05,
+      "loss": 0.2321,
+      "step": 111
+    },
+    {
+      "epoch": 0.458546571136131,
+      "grad_norm": 1.326955795288086,
+      "learning_rate": 1.933419956076584e-05,
+      "loss": 0.2516,
+      "step": 112
+    },
+    {
+      "epoch": 0.4626407369498465,
+      "grad_norm": 1.510201334953308,
+      "learning_rate": 1.9318049754832656e-05,
+      "loss": 0.2467,
+      "step": 113
+    },
+    {
+      "epoch": 0.46673490276356194,
+      "grad_norm": 2.9965062141418457,
+      "learning_rate": 1.9301713332493386e-05,
+      "loss": 0.2587,
+      "step": 114
+    },
+    {
+      "epoch": 0.47082906857727735,
+      "grad_norm": 1.3663560152053833,
+      "learning_rate": 1.9285190620924267e-05,
+      "loss": 0.2535,
+      "step": 115
+    },
+    {
+      "epoch": 0.4749232343909928,
+      "grad_norm": 1.1917448043823242,
+      "learning_rate": 1.926848195103242e-05,
+      "loss": 0.2292,
+      "step": 116
+    },
+    {
+      "epoch": 0.4790174002047083,
+      "grad_norm": 1.3093336820602417,
+      "learning_rate": 1.925158765744924e-05,
+      "loss": 0.229,
+      "step": 117
+    },
+    {
+      "epoch": 0.48311156601842375,
+      "grad_norm": 1.5121235847473145,
+      "learning_rate": 1.923450807852367e-05,
+      "loss": 0.2527,
+      "step": 118
+    },
+    {
+      "epoch": 0.4872057318321392,
+      "grad_norm": 1.461378812789917,
+      "learning_rate": 1.9217243556315445e-05,
+      "loss": 0.2631,
+      "step": 119
+    },
+    {
+      "epoch": 0.49129989764585463,
+      "grad_norm": 1.4177104234695435,
+      "learning_rate": 1.9199794436588244e-05,
+      "loss": 0.2378,
+      "step": 120
+    },
+    {
+      "epoch": 0.4953940634595701,
+      "grad_norm": 1.5456838607788086,
+      "learning_rate": 1.9182161068802742e-05,
+      "loss": 0.2491,
+      "step": 121
+    },
+    {
+      "epoch": 0.49948822927328557,
+      "grad_norm": 1.554958701133728,
+      "learning_rate": 1.916434380610963e-05,
+      "loss": 0.2461,
+      "step": 122
+    },
+    {
+      "epoch": 0.503582395087001,
+      "grad_norm": 1.4894706010818481,
+      "learning_rate": 1.9146343005342546e-05,
+      "loss": 0.2823,
+      "step": 123
+    },
+    {
+      "epoch": 0.5076765609007164,
+      "grad_norm": 1.359287142753601,
+      "learning_rate": 1.912815902701091e-05,
+      "loss": 0.2366,
+      "step": 124
+    },
+    {
+      "epoch": 0.5117707267144319,
+      "grad_norm": 1.4244722127914429,
+      "learning_rate": 1.9109792235292715e-05,
+      "loss": 0.2398,
+      "step": 125
+    },
+    {
+      "epoch": 0.5158648925281474,
+      "grad_norm": 1.4032812118530273,
+      "learning_rate": 1.909124299802724e-05,
+      "loss": 0.2601,
+      "step": 126
+    },
+    {
+      "epoch": 0.5199590583418628,
+      "grad_norm": 1.5199010372161865,
+      "learning_rate": 1.9072511686707663e-05,
+      "loss": 0.2458,
+      "step": 127
+    },
+    {
+      "epoch": 0.5240532241555783,
+      "grad_norm": 1.4342416524887085,
+      "learning_rate": 1.9053598676473656e-05,
+      "loss": 0.241,
+      "step": 128
+    },
+    {
+      "epoch": 0.5281473899692938,
+      "grad_norm": 1.338181972503662,
+      "learning_rate": 1.9034504346103825e-05,
+      "loss": 0.2508,
+      "step": 129
+    },
+    {
+      "epoch": 0.5322415557830092,
+      "grad_norm": 1.492775559425354,
+      "learning_rate": 1.9015229078008163e-05,
+      "loss": 0.2626,
+      "step": 130
+    },
+    {
+      "epoch": 0.5363357215967247,
+      "grad_norm": 1.3123077154159546,
+      "learning_rate": 1.8995773258220374e-05,
+      "loss": 0.258,
+      "step": 131
+    },
+    {
+      "epoch": 0.5404298874104401,
+      "grad_norm": 1.4001896381378174,
+      "learning_rate": 1.8976137276390145e-05,
+      "loss": 0.2425,
+      "step": 132
+    },
+    {
+      "epoch": 0.5445240532241555,
+      "grad_norm": 1.3989671468734741,
+      "learning_rate": 1.8956321525775337e-05,
+      "loss": 0.2781,
+      "step": 133
+    },
+    {
+      "epoch": 0.548618219037871,
+      "grad_norm": 1.260289192199707,
+      "learning_rate": 1.8936326403234125e-05,
+      "loss": 0.2432,
+      "step": 134
+    },
+    {
+      "epoch": 0.5527123848515865,
+      "grad_norm": 1.308370590209961,
+      "learning_rate": 1.891615230921703e-05,
+      "loss": 0.2408,
+      "step": 135
+    },
+    {
+      "epoch": 0.5568065506653019,
+      "grad_norm": 1.2668206691741943,
+      "learning_rate": 1.8895799647758912e-05,
+      "loss": 0.2408,
+      "step": 136
+    },
+    {
+      "epoch": 0.5609007164790174,
+      "grad_norm": 1.4105634689331055,
+      "learning_rate": 1.8875268826470875e-05,
+      "loss": 0.2688,
+      "step": 137
+    },
+    {
+      "epoch": 0.5649948822927329,
+      "grad_norm": 1.3877664804458618,
+      "learning_rate": 1.8854560256532098e-05,
+      "loss": 0.2379,
+      "step": 138
+    },
+    {
+      "epoch": 0.5690890481064483,
+      "grad_norm": 1.2643476724624634,
+      "learning_rate": 1.8833674352681613e-05,
+      "loss": 0.2375,
+      "step": 139
+    },
+    {
+      "epoch": 0.5731832139201638,
+      "grad_norm": 1.6310402154922485,
+      "learning_rate": 1.881261153320999e-05,
+      "loss": 0.2435,
+      "step": 140
+    },
+    {
+      "epoch": 0.5772773797338793,
+      "grad_norm": 1.35072660446167,
+      "learning_rate": 1.879137221995095e-05,
+      "loss": 0.2293,
+      "step": 141
+    },
+    {
+      "epoch": 0.5813715455475946,
+      "grad_norm": 1.409683346748352,
+      "learning_rate": 1.8769956838272937e-05,
+      "loss": 0.2584,
+      "step": 142
+    },
+    {
+      "epoch": 0.5854657113613101,
+      "grad_norm": 1.4319274425506592,
+      "learning_rate": 1.8748365817070586e-05,
+      "loss": 0.2669,
+      "step": 143
+    },
+    {
+      "epoch": 0.5895598771750256,
+      "grad_norm": 1.4259189367294312,
+      "learning_rate": 1.8726599588756144e-05,
+      "loss": 0.2447,
+      "step": 144
+    },
+    {
+      "epoch": 0.593654042988741,
+      "grad_norm": 1.4267385005950928,
+      "learning_rate": 1.8704658589250795e-05,
+      "loss": 0.2567,
+      "step": 145
+    },
+    {
+      "epoch": 0.5977482088024565,
+      "grad_norm": 1.4896032810211182,
+      "learning_rate": 1.868254325797594e-05,
+      "loss": 0.2363,
+      "step": 146
+    },
+    {
+      "epoch": 0.601842374616172,
+      "grad_norm": 1.355659008026123,
+      "learning_rate": 1.866025403784439e-05,
+      "loss": 0.2481,
+      "step": 147
+    },
+    {
+      "epoch": 0.6059365404298874,
+      "grad_norm": 1.2849416732788086,
+      "learning_rate": 1.8637791375251505e-05,
+      "loss": 0.2631,
+      "step": 148
+    },
+    {
+      "epoch": 0.6100307062436029,
+      "grad_norm": 1.279733419418335,
+      "learning_rate": 1.8615155720066247e-05,
+      "loss": 0.2487,
+      "step": 149
+    },
+    {
+      "epoch": 0.6141248720573184,
+      "grad_norm": 1.370506763458252,
+      "learning_rate": 1.859234752562217e-05,
+      "loss": 0.245,
+      "step": 150
+    },
+    {
+      "epoch": 0.6182190378710338,
+      "grad_norm": 1.352271556854248,
+      "learning_rate": 1.8569367248708343e-05,
+      "loss": 0.227,
+      "step": 151
+    },
+    {
+      "epoch": 0.6223132036847492,
+      "grad_norm": 1.3176748752593994,
+      "learning_rate": 1.8546215349560204e-05,
+      "loss": 0.2484,
+      "step": 152
+    },
+    {
+      "epoch": 0.6264073694984647,
+      "grad_norm": 1.386770486831665,
+      "learning_rate": 1.8522892291850335e-05,
+      "loss": 0.257,
+      "step": 153
+    },
+    {
+      "epoch": 0.6305015353121801,
+      "grad_norm": 1.5251948833465576,
+      "learning_rate": 1.849939854267919e-05,
+      "loss": 0.2727,
+      "step": 154
+    },
+    {
+      "epoch": 0.6345957011258956,
+      "grad_norm": 1.378936767578125,
+      "learning_rate": 1.847573457256571e-05,
+      "loss": 0.2337,
+      "step": 155
+    },
+    {
+      "epoch": 0.638689866939611,
+      "grad_norm": 1.3500293493270874,
+      "learning_rate": 1.845190085543795e-05,
+      "loss": 0.2618,
+      "step": 156
+    },
+    {
+      "epoch": 0.6427840327533265,
+      "grad_norm": 1.4462950229644775,
+      "learning_rate": 1.8427897868623535e-05,
+      "loss": 0.2588,
+      "step": 157
+    },
+    {
+      "epoch": 0.646878198567042,
+      "grad_norm": 1.2208290100097656,
+      "learning_rate": 1.840372609284013e-05,
+      "loss": 0.2927,
+      "step": 158
+    },
+    {
+      "epoch": 0.6509723643807575,
+      "grad_norm": 1.3487396240234375,
+      "learning_rate": 1.8379386012185813e-05,
+      "loss": 0.2417,
+      "step": 159
+    },
+    {
+      "epoch": 0.6550665301944729,
+      "grad_norm": 1.2655378580093384,
+      "learning_rate": 1.8354878114129368e-05,
+      "loss": 0.2428,
+      "step": 160
+    },
+    {
+      "epoch": 0.6591606960081884,
+      "grad_norm": 1.181028962135315,
+      "learning_rate": 1.8330202889500518e-05,
+      "loss": 0.2397,
+      "step": 161
+    },
+    {
+      "epoch": 0.6632548618219037,
+      "grad_norm": 1.3897309303283691,
+      "learning_rate": 1.8305360832480118e-05,
+      "loss": 0.2276,
+      "step": 162
+    },
+    {
+      "epoch": 0.6673490276356192,
+      "grad_norm": 1.340640902519226,
+      "learning_rate": 1.8280352440590236e-05,
+      "loss": 0.2375,
+      "step": 163
+    },
+    {
+      "epoch": 0.6714431934493347,
+      "grad_norm": 1.5388420820236206,
+      "learning_rate": 1.82551782146842e-05,
+      "loss": 0.2318,
+      "step": 164
+    },
+    {
+      "epoch": 0.6714431934493347,
+      "eval_loss": 0.2667195796966553,
+      "eval_runtime": 6.096,
+      "eval_samples_per_second": 12.959,
+      "eval_steps_per_second": 1.64,
+      "step": 164
+    },
+    {
+      "epoch": 0.6755373592630501,
+      "grad_norm": 1.4711631536483765,
+      "learning_rate": 1.8229838658936566e-05,
+      "loss": 0.2367,
+      "step": 165
+    },
+    {
+      "epoch": 0.6796315250767656,
+      "grad_norm": 1.3367588520050049,
+      "learning_rate": 1.8204334280833005e-05,
+      "loss": 0.2652,
+      "step": 166
+    },
+    {
+      "epoch": 0.6837256908904811,
+      "grad_norm": 1.5473802089691162,
+      "learning_rate": 1.817866559116017e-05,
+      "loss": 0.2693,
+      "step": 167
+    },
+    {
+      "epoch": 0.6878198567041965,
+      "grad_norm": 1.3539572954177856,
+      "learning_rate": 1.8152833103995443e-05,
+      "loss": 0.2431,
+      "step": 168
+    },
+    {
+      "epoch": 0.691914022517912,
+      "grad_norm": 1.3012574911117554,
+      "learning_rate": 1.8126837336696645e-05,
+      "loss": 0.2602,
+      "step": 169
+    },
+    {
+      "epoch": 0.6960081883316275,
+      "grad_norm": 1.3125128746032715,
+      "learning_rate": 1.8100678809891668e-05,
+      "loss": 0.2426,
+      "step": 170
+    },
+    {
+      "epoch": 0.7001023541453428,
+      "grad_norm": 1.2353239059448242,
+      "learning_rate": 1.807435804746807e-05,
+      "loss": 0.2176,
+      "step": 171
+    },
+    {
+      "epoch": 0.7041965199590583,
+      "grad_norm": 1.268622636795044,
+      "learning_rate": 1.8047875576562556e-05,
+      "loss": 0.2495,
+      "step": 172
+    },
+    {
+      "epoch": 0.7082906857727738,
+      "grad_norm": 1.3845950365066528,
+      "learning_rate": 1.802123192755044e-05,
+      "loss": 0.2643,
+      "step": 173
+    },
+    {
+      "epoch": 0.7123848515864892,
+      "grad_norm": 1.2450510263442993,
+      "learning_rate": 1.7994427634035016e-05,
+      "loss": 0.2445,
+      "step": 174
+    },
+    {
+      "epoch": 0.7164790174002047,
+      "grad_norm": 1.545483946800232,
+      "learning_rate": 1.796746323283686e-05,
+      "loss": 0.2594,
+      "step": 175
+    },
+    {
+      "epoch": 0.7205731832139202,
+      "grad_norm": 1.3044480085372925,
+      "learning_rate": 1.7940339263983112e-05,
+      "loss": 0.2563,
+      "step": 176
+    },
+    {
+      "epoch": 0.7246673490276356,
+      "grad_norm": 1.288320779800415,
+      "learning_rate": 1.791305627069662e-05,
+      "loss": 0.2478,
+      "step": 177
+    },
+    {
+      "epoch": 0.7287615148413511,
+      "grad_norm": 1.2802083492279053,
+      "learning_rate": 1.7885614799385086e-05,
+      "loss": 0.2331,
+      "step": 178
+    },
+    {
+      "epoch": 0.7328556806550666,
+      "grad_norm": 1.3979520797729492,
+      "learning_rate": 1.785801539963012e-05,
+      "loss": 0.2458,
+      "step": 179
+    },
+    {
+      "epoch": 0.736949846468782,
+      "grad_norm": 1.2247169017791748,
+      "learning_rate": 1.7830258624176224e-05,
+      "loss": 0.2132,
+      "step": 180
+    },
+    {
+      "epoch": 0.7410440122824974,
+      "grad_norm": 1.4145004749298096,
+      "learning_rate": 1.7802345028919728e-05,
+      "loss": 0.2402,
+      "step": 181
+    },
+    {
+      "epoch": 0.7451381780962129,
+      "grad_norm": 1.5720893144607544,
+      "learning_rate": 1.777427517289766e-05,
+      "loss": 0.258,
+      "step": 182
+    },
+    {
+      "epoch": 0.7492323439099283,
+      "grad_norm": 1.272928237915039,
+      "learning_rate": 1.7746049618276545e-05,
+      "loss": 0.2501,
+      "step": 183
+    },
+    {
+      "epoch": 0.7533265097236438,
+      "grad_norm": 1.4003905057907104,
+      "learning_rate": 1.7717668930341152e-05,
+      "loss": 0.2567,
+      "step": 184
+    },
+    {
+      "epoch": 0.7574206755373593,
+      "grad_norm": 1.4310050010681152,
+      "learning_rate": 1.768913367748316e-05,
+      "loss": 0.2443,
+      "step": 185
+    },
+    {
+      "epoch": 0.7615148413510747,
+      "grad_norm": 1.480718731880188,
+      "learning_rate": 1.766044443118978e-05,
+      "loss": 0.2564,
+      "step": 186
+    },
+    {
+      "epoch": 0.7656090071647902,
+      "grad_norm": 1.331586480140686,
+      "learning_rate": 1.7631601766032337e-05,
+      "loss": 0.2467,
+      "step": 187
+    },
+    {
+      "epoch": 0.7697031729785057,
+      "grad_norm": 1.2713260650634766,
+      "learning_rate": 1.7602606259654704e-05,
+      "loss": 0.2406,
+      "step": 188
+    },
+    {
+      "epoch": 0.7737973387922211,
+      "grad_norm": 1.4929317235946655,
+      "learning_rate": 1.7573458492761802e-05,
+      "loss": 0.2515,
+      "step": 189
+    },
+    {
+      "epoch": 0.7778915046059366,
+      "grad_norm": 1.117374062538147,
+      "learning_rate": 1.7544159049107902e-05,
+      "loss": 0.2165,
+      "step": 190
+    },
+    {
+      "epoch": 0.781985670419652,
+      "grad_norm": 1.2868603467941284,
+      "learning_rate": 1.7514708515485002e-05,
+      "loss": 0.2469,
+      "step": 191
+    },
+    {
+      "epoch": 0.7860798362333674,
+      "grad_norm": 1.3611894845962524,
+      "learning_rate": 1.7485107481711014e-05,
+      "loss": 0.2468,
+      "step": 192
+    },
+    {
+      "epoch": 0.7901740020470829,
+      "grad_norm": 1.3682395219802856,
+      "learning_rate": 1.7455356540617988e-05,
+      "loss": 0.2548,
+      "step": 193
+    },
+    {
+      "epoch": 0.7942681678607983,
+      "grad_norm": 1.2504215240478516,
+      "learning_rate": 1.7425456288040236e-05,
+      "loss": 0.2563,
+      "step": 194
+    },
+    {
+      "epoch": 0.7983623336745138,
+      "grad_norm": 1.3082585334777832,
+      "learning_rate": 1.7395407322802374e-05,
+      "loss": 0.232,
+      "step": 195
+    },
+    {
+      "epoch": 0.8024564994882293,
+      "grad_norm": 1.409508466720581,
+      "learning_rate": 1.736521024670737e-05,
+      "loss": 0.2438,
+      "step": 196
+    },
+    {
+      "epoch": 0.8065506653019447,
+      "grad_norm": 1.1714893579483032,
+      "learning_rate": 1.733486566452446e-05,
+      "loss": 0.2201,
+      "step": 197
+    },
+    {
+      "epoch": 0.8106448311156602,
+      "grad_norm": 1.3167288303375244,
+      "learning_rate": 1.7304374183977032e-05,
+      "loss": 0.2359,
+      "step": 198
+    },
+    {
+      "epoch": 0.8147389969293757,
+      "grad_norm": 1.2892719507217407,
+      "learning_rate": 1.7273736415730488e-05,
+      "loss": 0.2373,
+      "step": 199
+    },
+    {
+      "epoch": 0.8188331627430911,
+      "grad_norm": 1.2318534851074219,
+      "learning_rate": 1.7242952973379983e-05,
+      "loss": 0.2493,
+      "step": 200
+    },
+    {
+      "epoch": 0.8229273285568065,
+      "grad_norm": 1.2666089534759521,
+      "learning_rate": 1.7212024473438145e-05,
+      "loss": 0.2335,
+      "step": 201
+    },
+    {
+      "epoch": 0.827021494370522,
+      "grad_norm": 1.3021701574325562,
+      "learning_rate": 1.7180951535322742e-05,
+      "loss": 0.2457,
+      "step": 202
+    },
+    {
+      "epoch": 0.8311156601842374,
+      "grad_norm": 1.2390035390853882,
+      "learning_rate": 1.7149734781344247e-05,
+      "loss": 0.2317,
+      "step": 203
+    },
+    {
+      "epoch": 0.8352098259979529,
+      "grad_norm": 1.373651146888733,
+      "learning_rate": 1.7118374836693407e-05,
+      "loss": 0.2477,
+      "step": 204
+    },
+    {
+      "epoch": 0.8393039918116684,
+      "grad_norm": 1.4125158786773682,
+      "learning_rate": 1.7086872329428702e-05,
+      "loss": 0.2716,
+      "step": 205
+    },
+    {
+      "epoch": 0.8433981576253838,
+      "grad_norm": 1.3470803499221802,
+      "learning_rate": 1.705522789046377e-05,
+      "loss": 0.264,
+      "step": 206
+    },
+    {
+      "epoch": 0.8474923234390993,
+      "grad_norm": 1.419498324394226,
+      "learning_rate": 1.7023442153554776e-05,
+      "loss": 0.2626,
+      "step": 207
+    },
+    {
+      "epoch": 0.8515864892528148,
+      "grad_norm": 1.4023360013961792,
+      "learning_rate": 1.6991515755287715e-05,
+      "loss": 0.2786,
+      "step": 208
+    },
+    {
+      "epoch": 0.8556806550665302,
+      "grad_norm": 1.3539841175079346,
+      "learning_rate": 1.695944933506567e-05,
+      "loss": 0.2623,
+      "step": 209
+    },
+    {
+      "epoch": 0.8597748208802457,
+      "grad_norm": 1.4278448820114136,
+      "learning_rate": 1.6927243535095995e-05,
+      "loss": 0.2491,
+      "step": 210
+    },
+    {
+      "epoch": 0.8638689866939611,
+      "grad_norm": 1.4281779527664185,
+      "learning_rate": 1.6894899000377462e-05,
+      "loss": 0.2303,
+      "step": 211
+    },
+    {
+      "epoch": 0.8679631525076765,
+      "grad_norm": 1.2435628175735474,
+      "learning_rate": 1.686241637868734e-05,
+      "loss": 0.254,
+      "step": 212
+    },
+    {
+      "epoch": 0.872057318321392,
+      "grad_norm": 1.2799077033996582,
+      "learning_rate": 1.6829796320568416e-05,
+      "loss": 0.2253,
+      "step": 213
+    },
+    {
+      "epoch": 0.8761514841351075,
+      "grad_norm": 1.4002091884613037,
+      "learning_rate": 1.6797039479315994e-05,
+      "loss": 0.2568,
+      "step": 214
+    },
+    {
+      "epoch": 0.8802456499488229,
+      "grad_norm": 1.2796311378479004,
+      "learning_rate": 1.6764146510964762e-05,
+      "loss": 0.2488,
+      "step": 215
+    },
+    {
+      "epoch": 0.8843398157625384,
+      "grad_norm": 1.28677237033844,
+      "learning_rate": 1.67311180742757e-05,
+      "loss": 0.2619,
+      "step": 216
+    },
+    {
+      "epoch": 0.8884339815762539,
+      "grad_norm": 1.5529658794403076,
+      "learning_rate": 1.669795483072287e-05,
+      "loss": 0.2312,
+      "step": 217
+    },
+    {
+      "epoch": 0.8925281473899693,
+      "grad_norm": 1.2904109954833984,
+      "learning_rate": 1.6664657444480145e-05,
+      "loss": 0.2366,
+      "step": 218
+    },
+    {
+      "epoch": 0.8966223132036848,
+      "grad_norm": 1.2689939737319946,
+      "learning_rate": 1.6631226582407954e-05,
+      "loss": 0.2446,
+      "step": 219
+    },
+    {
+      "epoch": 0.9007164790174002,
+      "grad_norm": 1.2608931064605713,
+      "learning_rate": 1.6597662914039885e-05,
+      "loss": 0.2419,
+      "step": 220
+    },
+    {
+      "epoch": 0.9048106448311156,
+      "grad_norm": 1.2563594579696655,
+      "learning_rate": 1.65639671115693e-05,
+      "loss": 0.2363,
+      "step": 221
+    },
+    {
+      "epoch": 0.9089048106448311,
+      "grad_norm": 1.320057988166809,
+      "learning_rate": 1.653013984983585e-05,
+      "loss": 0.2508,
+      "step": 222
+    },
+    {
+      "epoch": 0.9129989764585466,
+      "grad_norm": 1.2236367464065552,
+      "learning_rate": 1.6496181806312005e-05,
+      "loss": 0.2263,
+      "step": 223
+    },
+    {
+      "epoch": 0.917093142272262,
+      "grad_norm": 1.293455958366394,
+      "learning_rate": 1.6462093661089432e-05,
+      "loss": 0.2504,
+      "step": 224
+    },
+    {
+      "epoch": 0.9211873080859775,
+      "grad_norm": 1.3519924879074097,
+      "learning_rate": 1.6427876096865394e-05,
+      "loss": 0.2458,
+      "step": 225
+    },
+    {
+      "epoch": 0.925281473899693,
+      "grad_norm": 1.1655514240264893,
+      "learning_rate": 1.6393529798929103e-05,
+      "loss": 0.249,
+      "step": 226
+    },
+    {
+      "epoch": 0.9293756397134084,
+      "grad_norm": 1.3402174711227417,
+      "learning_rate": 1.635905545514795e-05,
+      "loss": 0.2426,
+      "step": 227
+    },
+    {
+      "epoch": 0.9334698055271239,
+      "grad_norm": 1.3474458456039429,
+      "learning_rate": 1.6324453755953772e-05,
+      "loss": 0.2497,
+      "step": 228
+    },
+    {
+      "epoch": 0.9375639713408394,
+      "grad_norm": 1.2284787893295288,
+      "learning_rate": 1.6289725394328998e-05,
+      "loss": 0.2438,
+      "step": 229
+    },
+    {
+      "epoch": 0.9416581371545547,
+      "grad_norm": 1.2306231260299683,
+      "learning_rate": 1.6254871065792776e-05,
+      "loss": 0.248,
+      "step": 230
+    },
+    {
+      "epoch": 0.9457523029682702,
+      "grad_norm": 1.3834123611450195,
+      "learning_rate": 1.621989146838704e-05,
+      "loss": 0.2798,
+      "step": 231
+    },
+    {
+      "epoch": 0.9498464687819856,
+      "grad_norm": 1.333827257156372,
+      "learning_rate": 1.618478730266255e-05,
+      "loss": 0.2585,
+      "step": 232
+    },
+    {
+      "epoch": 0.9539406345957011,
+      "grad_norm": 1.272150993347168,
+      "learning_rate": 1.6149559271664835e-05,
+      "loss": 0.2512,
+      "step": 233
+    },
+    {
+      "epoch": 0.9580348004094166,
+      "grad_norm": 1.3800173997879028,
+      "learning_rate": 1.6114208080920125e-05,
+      "loss": 0.2483,
+      "step": 234
+    },
+    {
+      "epoch": 0.962128966223132,
+      "grad_norm": 1.2623674869537354,
+      "learning_rate": 1.607873443842122e-05,
+      "loss": 0.2447,
+      "step": 235
+    },
+    {
+      "epoch": 0.9662231320368475,
+      "grad_norm": 1.2797845602035522,
+      "learning_rate": 1.6043139054613326e-05,
+      "loss": 0.2258,
+      "step": 236
+    },
+    {
+      "epoch": 0.970317297850563,
+      "grad_norm": 1.3689030408859253,
+      "learning_rate": 1.600742264237979e-05,
+      "loss": 0.247,
+      "step": 237
+    },
+    {
+      "epoch": 0.9744114636642784,
+      "grad_norm": 1.3384076356887817,
+      "learning_rate": 1.5971585917027864e-05,
+      "loss": 0.2286,
+      "step": 238
+    },
+    {
+      "epoch": 0.9785056294779939,
+      "grad_norm": 1.210942268371582,
+      "learning_rate": 1.5935629596274345e-05,
+      "loss": 0.2402,
+      "step": 239
+    },
+    {
+      "epoch": 0.9825997952917093,
+      "grad_norm": 1.2547039985656738,
+      "learning_rate": 1.5899554400231233e-05,
+      "loss": 0.2386,
+      "step": 240
+    },
+    {
+      "epoch": 0.9866939611054247,
+      "grad_norm": 1.337138056755066,
+      "learning_rate": 1.586336105139127e-05,
+      "loss": 0.2686,
+      "step": 241
+    },
+    {
+      "epoch": 0.9907881269191402,
+      "grad_norm": 1.273237943649292,
+      "learning_rate": 1.5827050274613512e-05,
+      "loss": 0.2653,
+      "step": 242
+    },
+    {
+      "epoch": 0.9948822927328557,
+      "grad_norm": 1.3556797504425049,
+      "learning_rate": 1.579062279710879e-05,
+      "loss": 0.2365,
+      "step": 243
+    },
+    {
+      "epoch": 0.9989764585465711,
+      "grad_norm": 1.2234649658203125,
+      "learning_rate": 1.5754079348425137e-05,
+      "loss": 0.239,
+      "step": 244
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2234649658203125,
+      "learning_rate": 1.57174206604332e-05,
+      "loss": 0.1857,
+      "step": 245
+    },
+    {
+      "epoch": 1.0040941658137155,
+      "grad_norm": 0.9376870393753052,
+      "learning_rate": 1.568064746731156e-05,
+      "loss": 0.1688,
+      "step": 246
+    },
+    {
+      "epoch": 1.0040941658137155,
+      "eval_loss": 0.26177823543548584,
+      "eval_runtime": 5.0735,
+      "eval_samples_per_second": 15.571,
+      "eval_steps_per_second": 1.971,
+      "step": 246
+    },
+    {
+      "epoch": 1.008188331627431,
+      "grad_norm": 0.8188478350639343,
+      "learning_rate": 1.564376050553205e-05,
+      "loss": 0.1587,
+      "step": 247
+    },
+    {
+      "epoch": 1.0122824974411464,
+      "grad_norm": 1.0571590662002563,
+      "learning_rate": 1.560676051384499e-05,
+      "loss": 0.1842,
+      "step": 248
+    },
+    {
+      "epoch": 1.0163766632548619,
+      "grad_norm": 0.8216506242752075,
+      "learning_rate": 1.5569648233264395e-05,
+      "loss": 0.1591,
+      "step": 249
+    },
+    {
+      "epoch": 1.0204708290685773,
+      "grad_norm": 0.8552656769752502,
+      "learning_rate": 1.553242440705314e-05,
+      "loss": 0.1574,
+      "step": 250
+    },
+    {
+      "epoch": 1.0245649948822928,
+      "grad_norm": 0.958118200302124,
+      "learning_rate": 1.5495089780708062e-05,
+      "loss": 0.1543,
+      "step": 251
+    },
+    {
+      "epoch": 1.0286591606960083,
+      "grad_norm": 0.8587727546691895,
+      "learning_rate": 1.5457645101945046e-05,
+      "loss": 0.1647,
+      "step": 252
+    },
+    {
+      "epoch": 1.0327533265097237,
+      "grad_norm": 0.9147894978523254,
+      "learning_rate": 1.5420091120684042e-05,
+      "loss": 0.1594,
+      "step": 253
+    },
+    {
+      "epoch": 1.0368474923234392,
+      "grad_norm": 0.9051440358161926,
+      "learning_rate": 1.538242858903404e-05,
+      "loss": 0.1598,
+      "step": 254
+    },
+    {
+      "epoch": 1.0409416581371547,
+      "grad_norm": 0.9272040128707886,
+      "learning_rate": 1.5344658261278013e-05,
+      "loss": 0.1435,
+      "step": 255
+    },
+    {
+      "epoch": 1.04503582395087,
+      "grad_norm": 0.9846145510673523,
+      "learning_rate": 1.530678089385782e-05,
+      "loss": 0.1566,
+      "step": 256
+    },
+    {
+      "epoch": 1.0491299897645854,
+      "grad_norm": 1.080641269683838,
+      "learning_rate": 1.5268797245359035e-05,
+      "loss": 0.1659,
+      "step": 257
+    },
+    {
+      "epoch": 1.0532241555783008,
+      "grad_norm": 1.1138889789581299,
+      "learning_rate": 1.5230708076495777e-05,
+      "loss": 0.1633,
+      "step": 258
+    },
+    {
+      "epoch": 1.0573183213920163,
+      "grad_norm": 1.0673142671585083,
+      "learning_rate": 1.519251415009546e-05,
+      "loss": 0.1719,
+      "step": 259
+    },
+    {
+      "epoch": 1.0614124872057318,
+      "grad_norm": 0.9915375113487244,
+      "learning_rate": 1.5154216231083522e-05,
+      "loss": 0.1391,
+      "step": 260
+    },
+    {
+      "epoch": 1.0655066530194472,
+      "grad_norm": 1.13100004196167,
+      "learning_rate": 1.5115815086468103e-05,
+      "loss": 0.1795,
+      "step": 261
+    },
+    {
+      "epoch": 1.0696008188331627,
+      "grad_norm": 1.2729498147964478,
+      "learning_rate": 1.507731148532468e-05,
+      "loss": 0.1563,
+      "step": 262
+    },
+    {
+      "epoch": 1.0736949846468782,
+      "grad_norm": 1.0953161716461182,
+      "learning_rate": 1.5038706198780673e-05,
+      "loss": 0.1698,
+      "step": 263
+    },
+    {
+      "epoch": 1.0777891504605936,
+      "grad_norm": 1.047824501991272,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.1746,
+      "step": 264
+    },
+    {
+      "epoch": 1.0818833162743091,
+      "grad_norm": 1.0675163269042969,
+      "learning_rate": 1.496119366416759e-05,
+      "loss": 0.1683,
+      "step": 265
+    },
+    {
+      "epoch": 1.0859774820880246,
+      "grad_norm": 0.9779859185218811,
+      "learning_rate": 1.492228796847385e-05,
+      "loss": 0.1768,
+      "step": 266
+    },
+    {
+      "epoch": 1.09007164790174,
+      "grad_norm": 1.0177017450332642,
+      "learning_rate": 1.4883283692099114e-05,
+      "loss": 0.1609,
+      "step": 267
+    },
+    {
+      "epoch": 1.0941658137154555,
+      "grad_norm": 1.0419244766235352,
+      "learning_rate": 1.4844181616198028e-05,
+      "loss": 0.1555,
+      "step": 268
+    },
+    {
+      "epoch": 1.098259979529171,
+      "grad_norm": 1.1818829774856567,
+      "learning_rate": 1.4804982523883915e-05,
+      "loss": 0.1582,
+      "step": 269
+    },
+    {
+      "epoch": 1.1023541453428864,
+      "grad_norm": 0.8959009051322937,
+      "learning_rate": 1.4765687200213079e-05,
+      "loss": 0.1408,
+      "step": 270
+    },
+    {
+      "epoch": 1.106448311156602,
+      "grad_norm": 1.008933424949646,
+      "learning_rate": 1.4726296432169095e-05,
+      "loss": 0.1518,
+      "step": 271
+    },
+    {
+      "epoch": 1.1105424769703174,
+      "grad_norm": 0.9755818843841553,
+      "learning_rate": 1.4686811008647037e-05,
+      "loss": 0.1513,
+      "step": 272
+    },
+    {
+      "epoch": 1.1146366427840328,
+      "grad_norm": 1.0435041189193726,
+      "learning_rate": 1.4647231720437687e-05,
+      "loss": 0.1673,
+      "step": 273
+    },
+    {
+      "epoch": 1.118730808597748,
+      "grad_norm": 0.9561364054679871,
+      "learning_rate": 1.4607559360211688e-05,
+      "loss": 0.1575,
+      "step": 274
+    },
+    {
+      "epoch": 1.1228249744114636,
+      "grad_norm": 1.0000271797180176,
+      "learning_rate": 1.456779472250368e-05,
+      "loss": 0.1677,
+      "step": 275
+    },
+    {
+      "epoch": 1.126919140225179,
+      "grad_norm": 1.1137831211090088,
+      "learning_rate": 1.4527938603696376e-05,
+      "loss": 0.1601,
+      "step": 276
+    },
+    {
+      "epoch": 1.1310133060388945,
+      "grad_norm": 0.9575244188308716,
+      "learning_rate": 1.4487991802004625e-05,
+      "loss": 0.1591,
+      "step": 277
+    },
+    {
+      "epoch": 1.13510747185261,
+      "grad_norm": 0.9593146443367004,
+      "learning_rate": 1.4447955117459414e-05,
+      "loss": 0.174,
+      "step": 278
+    },
+    {
+      "epoch": 1.1392016376663254,
+      "grad_norm": 1.0342645645141602,
+      "learning_rate": 1.4407829351891858e-05,
+      "loss": 0.1808,
+      "step": 279
+    },
+    {
+      "epoch": 1.143295803480041,
+      "grad_norm": 0.908888578414917,
+      "learning_rate": 1.436761530891713e-05,
+      "loss": 0.1587,
+      "step": 280
+    },
+    {
+      "epoch": 1.1473899692937564,
+      "grad_norm": 0.8734580874443054,
+      "learning_rate": 1.4327313793918362e-05,
+      "loss": 0.1364,
+      "step": 281
+    },
+    {
+      "epoch": 1.1514841351074718,
+      "grad_norm": 1.0717835426330566,
+      "learning_rate": 1.4286925614030542e-05,
+      "loss": 0.1673,
+      "step": 282
+    },
+    {
+      "epoch": 1.1555783009211873,
+      "grad_norm": 1.099330186843872,
+      "learning_rate": 1.4246451578124321e-05,
+      "loss": 0.1783,
+      "step": 283
+    },
+    {
+      "epoch": 1.1596724667349028,
+      "grad_norm": 1.0060065984725952,
+      "learning_rate": 1.4205892496789816e-05,
+      "loss": 0.1613,
+      "step": 284
+    },
+    {
+      "epoch": 1.1637666325486182,
+      "grad_norm": 1.0185686349868774,
+      "learning_rate": 1.4165249182320401e-05,
+      "loss": 0.153,
+      "step": 285
+    },
+    {
+      "epoch": 1.1678607983623337,
+      "grad_norm": 1.0007094144821167,
+      "learning_rate": 1.4124522448696407e-05,
+      "loss": 0.1681,
+      "step": 286
+    },
+    {
+      "epoch": 1.1719549641760492,
+      "grad_norm": 0.886319637298584,
+      "learning_rate": 1.4083713111568841e-05,
+      "loss": 0.1712,
+      "step": 287
+    },
+    {
+      "epoch": 1.1760491299897646,
+      "grad_norm": 0.8676832914352417,
+      "learning_rate": 1.404282198824305e-05,
+      "loss": 0.1458,
+      "step": 288
+    },
+    {
+      "epoch": 1.18014329580348,
+      "grad_norm": 0.9190279841423035,
+      "learning_rate": 1.4001849897662337e-05,
+      "loss": 0.1468,
+      "step": 289
+    },
+    {
+      "epoch": 1.1842374616171956,
+      "grad_norm": 0.931041419506073,
+      "learning_rate": 1.396079766039157e-05,
+      "loss": 0.16,
+      "step": 290
+    },
+    {
+      "epoch": 1.188331627430911,
+      "grad_norm": 0.9734114408493042,
+      "learning_rate": 1.3919666098600753e-05,
+      "loss": 0.1545,
+      "step": 291
+    },
+    {
+      "epoch": 1.1924257932446265,
+      "grad_norm": 0.933809757232666,
+      "learning_rate": 1.387845603604855e-05,
+      "loss": 0.156,
+      "step": 292
+    },
+    {
+      "epoch": 1.196519959058342,
+      "grad_norm": 0.9046220779418945,
+      "learning_rate": 1.3837168298065798e-05,
+      "loss": 0.1378,
+      "step": 293
+    },
+    {
+      "epoch": 1.2006141248720574,
+      "grad_norm": 0.9085408449172974,
+      "learning_rate": 1.3795803711538966e-05,
+      "loss": 0.1616,
+      "step": 294
+    },
+    {
+      "epoch": 1.204708290685773,
+      "grad_norm": 0.9923974275588989,
+      "learning_rate": 1.37543631048936e-05,
+      "loss": 0.1797,
+      "step": 295
+    },
+    {
+      "epoch": 1.2088024564994881,
+      "grad_norm": 0.9671124219894409,
+      "learning_rate": 1.3712847308077737e-05,
+      "loss": 0.1558,
+      "step": 296
+    },
+    {
+      "epoch": 1.2128966223132036,
+      "grad_norm": 1.1265525817871094,
+      "learning_rate": 1.3671257152545277e-05,
+      "loss": 0.1582,
+      "step": 297
+    },
+    {
+      "epoch": 1.216990788126919,
+      "grad_norm": 1.066918969154358,
+      "learning_rate": 1.3629593471239328e-05,
+      "loss": 0.1552,
+      "step": 298
+    },
+    {
+      "epoch": 1.2210849539406345,
+      "grad_norm": 0.9553365111351013,
+      "learning_rate": 1.3587857098575534e-05,
+      "loss": 0.1614,
+      "step": 299
+    },
+    {
+      "epoch": 1.22517911975435,
+      "grad_norm": 1.0191755294799805,
+      "learning_rate": 1.3546048870425356e-05,
+      "loss": 0.1661,
+      "step": 300
+    },
+    {
+      "epoch": 1.2292732855680655,
+      "grad_norm": 0.8861321806907654,
+      "learning_rate": 1.350416962409934e-05,
+      "loss": 0.1383,
+      "step": 301
+    },
+    {
+      "epoch": 1.233367451381781,
+      "grad_norm": 0.9416659474372864,
+      "learning_rate": 1.346222019833033e-05,
+      "loss": 0.1429,
+      "step": 302
+    },
+    {
+      "epoch": 1.2374616171954964,
+      "grad_norm": 1.0216457843780518,
+      "learning_rate": 1.342020143325669e-05,
+      "loss": 0.1503,
+      "step": 303
+    },
+    {
+      "epoch": 1.2415557830092119,
+      "grad_norm": 1.08574378490448,
+      "learning_rate": 1.3378114170405473e-05,
+      "loss": 0.1685,
+      "step": 304
+    },
+    {
+      "epoch": 1.2456499488229273,
+      "grad_norm": 0.926025390625,
+      "learning_rate": 1.3335959252675566e-05,
+      "loss": 0.159,
+      "step": 305
+    },
+    {
+      "epoch": 1.2497441146366428,
+      "grad_norm": 1.0333497524261475,
+      "learning_rate": 1.3293737524320798e-05,
+      "loss": 0.142,
+      "step": 306
+    },
+    {
+      "epoch": 1.2538382804503583,
+      "grad_norm": 0.9234523773193359,
+      "learning_rate": 1.3251449830933052e-05,
+      "loss": 0.1512,
+      "step": 307
+    },
+    {
+      "epoch": 1.2579324462640737,
+      "grad_norm": 0.9113657474517822,
+      "learning_rate": 1.3209097019425317e-05,
+      "loss": 0.162,
+      "step": 308
+    },
+    {
+      "epoch": 1.2620266120777892,
+      "grad_norm": 0.9163710474967957,
+      "learning_rate": 1.3166679938014728e-05,
+      "loss": 0.1676,
+      "step": 309
+    },
+    {
+      "epoch": 1.2661207778915047,
+      "grad_norm": 0.8762107491493225,
+      "learning_rate": 1.3124199436205575e-05,
+      "loss": 0.1564,
+      "step": 310
+    },
+    {
+      "epoch": 1.27021494370522,
+      "grad_norm": 0.9909912943840027,
+      "learning_rate": 1.3081656364772308e-05,
+      "loss": 0.162,
+      "step": 311
+    },
+    {
+      "epoch": 1.2743091095189354,
+      "grad_norm": 1.0118768215179443,
+      "learning_rate": 1.303905157574247e-05,
+      "loss": 0.1573,
+      "step": 312
+    },
+    {
+      "epoch": 1.2784032753326509,
+      "grad_norm": 0.9785376191139221,
+      "learning_rate": 1.2996385922379657e-05,
+      "loss": 0.1695,
+      "step": 313
+    },
+    {
+      "epoch": 1.2824974411463663,
+      "grad_norm": 1.0604181289672852,
+      "learning_rate": 1.2953660259166413e-05,
+      "loss": 0.1817,
+      "step": 314
+    },
+    {
+      "epoch": 1.2865916069600818,
+      "grad_norm": 0.909048318862915,
+      "learning_rate": 1.291087544178713e-05,
+      "loss": 0.1599,
+      "step": 315
+    },
+    {
+      "epoch": 1.2906857727737973,
+      "grad_norm": 0.988167405128479,
+      "learning_rate": 1.2868032327110904e-05,
+      "loss": 0.1586,
+      "step": 316
+    },
+    {
+      "epoch": 1.2947799385875127,
+      "grad_norm": 1.0076347589492798,
+      "learning_rate": 1.2825131773174371e-05,
+      "loss": 0.1791,
+      "step": 317
+    },
+    {
+      "epoch": 1.2988741044012282,
+      "grad_norm": 1.09652841091156,
+      "learning_rate": 1.2782174639164528e-05,
+      "loss": 0.1585,
+      "step": 318
+    },
+    {
+      "epoch": 1.3029682702149437,
+      "grad_norm": 1.0916470289230347,
+      "learning_rate": 1.2739161785401525e-05,
+      "loss": 0.1782,
+      "step": 319
+    },
+    {
+      "epoch": 1.3070624360286591,
+      "grad_norm": 1.0601223707199097,
+      "learning_rate": 1.269609407332144e-05,
+      "loss": 0.1654,
+      "step": 320
+    },
+    {
+      "epoch": 1.3111566018423746,
+      "grad_norm": 1.1371649503707886,
+      "learning_rate": 1.2652972365459008e-05,
+      "loss": 0.1763,
+      "step": 321
+    },
+    {
+      "epoch": 1.31525076765609,
+      "grad_norm": 1.0163546800613403,
+      "learning_rate": 1.2609797525430374e-05,
+      "loss": 0.17,
+      "step": 322
+    },
+    {
+      "epoch": 1.3193449334698055,
+      "grad_norm": 0.9880717396736145,
+      "learning_rate": 1.2566570417915769e-05,
+      "loss": 0.1691,
+      "step": 323
+    },
+    {
+      "epoch": 1.323439099283521,
+      "grad_norm": 0.8969373106956482,
+      "learning_rate": 1.2523291908642219e-05,
+      "loss": 0.1625,
+      "step": 324
+    },
+    {
+      "epoch": 1.3275332650972365,
+      "grad_norm": 1.0454288721084595,
+      "learning_rate": 1.2479962864366186e-05,
+      "loss": 0.1669,
+      "step": 325
+    },
+    {
+      "epoch": 1.331627430910952,
+      "grad_norm": 1.0298652648925781,
+      "learning_rate": 1.243658415285622e-05,
+      "loss": 0.1822,
+      "step": 326
+    },
+    {
+      "epoch": 1.3357215967246674,
+      "grad_norm": 0.9368167519569397,
+      "learning_rate": 1.2393156642875579e-05,
+      "loss": 0.1614,
+      "step": 327
+    },
+    {
+      "epoch": 1.3398157625383829,
+      "grad_norm": 0.9903757572174072,
+      "learning_rate": 1.2349681204164823e-05,
+      "loss": 0.1467,
+      "step": 328
+    },
+    {
+      "epoch": 1.3398157625383829,
+      "eval_loss": 0.26726678013801575,
+      "eval_runtime": 5.839,
+      "eval_samples_per_second": 13.53,
+      "eval_steps_per_second": 1.713,
+      "step": 328
+    },
+    {
+      "epoch": 1.3439099283520983,
+      "grad_norm": 0.9085035920143127,
+      "learning_rate": 1.2306158707424402e-05,
+      "loss": 0.1486,
+      "step": 329
+    },
+    {
+      "epoch": 1.3480040941658138,
+      "grad_norm": 0.9950000643730164,
+      "learning_rate": 1.2262590024297226e-05,
+      "loss": 0.1586,
+      "step": 330
+    },
+    {
+      "epoch": 1.3520982599795293,
+      "grad_norm": 0.963299572467804,
+      "learning_rate": 1.2218976027351177e-05,
+      "loss": 0.1508,
+      "step": 331
+    },
+    {
+      "epoch": 1.3561924257932447,
+      "grad_norm": 0.9706941246986389,
+      "learning_rate": 1.2175317590061676e-05,
+      "loss": 0.1576,
+      "step": 332
+    },
+    {
+      "epoch": 1.3602865916069602,
+      "grad_norm": 1.1053946018218994,
+      "learning_rate": 1.2131615586794162e-05,
+      "loss": 0.1632,
+      "step": 333
+    },
+    {
+      "epoch": 1.3643807574206757,
+      "grad_norm": 0.9622719287872314,
+      "learning_rate": 1.2087870892786588e-05,
+      "loss": 0.1629,
+      "step": 334
+    },
+    {
+      "epoch": 1.3684749232343911,
+      "grad_norm": 0.9611127972602844,
+      "learning_rate": 1.2044084384131891e-05,
+      "loss": 0.1517,
+      "step": 335
+    },
+    {
+      "epoch": 1.3725690890481064,
+      "grad_norm": 0.9548255801200867,
+      "learning_rate": 1.2000256937760446e-05,
+      "loss": 0.1527,
+      "step": 336
+    },
+    {
+      "epoch": 1.3766632548618218,
+      "grad_norm": 0.9860974550247192,
+      "learning_rate": 1.1956389431422508e-05,
+      "loss": 0.1729,
+      "step": 337
+    },
+    {
+      "epoch": 1.3807574206755373,
+      "grad_norm": 1.0746179819107056,
+      "learning_rate": 1.1912482743670624e-05,
+      "loss": 0.1713,
+      "step": 338
+    },
+    {
+      "epoch": 1.3848515864892528,
+      "grad_norm": 0.8706703782081604,
+      "learning_rate": 1.1868537753842052e-05,
+      "loss": 0.1521,
+      "step": 339
+    },
+    {
+      "epoch": 1.3889457523029682,
+      "grad_norm": 0.9581889510154724,
+      "learning_rate": 1.1824555342041129e-05,
+      "loss": 0.1401,
+      "step": 340
+    },
+    {
+      "epoch": 1.3930399181166837,
+      "grad_norm": 0.9390183091163635,
+      "learning_rate": 1.1780536389121668e-05,
+      "loss": 0.1391,
+      "step": 341
+    },
+    {
+      "epoch": 1.3971340839303992,
+      "grad_norm": 0.9149916768074036,
+      "learning_rate": 1.1736481776669307e-05,
+      "loss": 0.1492,
+      "step": 342
+    },
+    {
+      "epoch": 1.4012282497441146,
+      "grad_norm": 0.9424813389778137,
+      "learning_rate": 1.1692392386983837e-05,
+      "loss": 0.1716,
+      "step": 343
+    },
+    {
+      "epoch": 1.40532241555783,
+      "grad_norm": 0.9241910576820374,
+      "learning_rate": 1.1648269103061567e-05,
+      "loss": 0.166,
+      "step": 344
+    },
+    {
+      "epoch": 1.4094165813715456,
+      "grad_norm": 1.0855869054794312,
+      "learning_rate": 1.1604112808577603e-05,
+      "loss": 0.1653,
+      "step": 345
+    },
+    {
+      "epoch": 1.413510747185261,
+      "grad_norm": 0.9882118105888367,
+      "learning_rate": 1.155992438786818e-05,
+      "loss": 0.1578,
+      "step": 346
+    },
+    {
+      "epoch": 1.4176049129989765,
+      "grad_norm": 0.9933427572250366,
+      "learning_rate": 1.1515704725912926e-05,
+      "loss": 0.1506,
+      "step": 347
+    },
+    {
+      "epoch": 1.421699078812692,
+      "grad_norm": 0.9362509250640869,
+      "learning_rate": 1.1471454708317163e-05,
+      "loss": 0.1704,
+      "step": 348
+    },
+    {
+      "epoch": 1.4257932446264074,
+      "grad_norm": 0.9945564866065979,
+      "learning_rate": 1.1427175221294145e-05,
+      "loss": 0.1619,
+      "step": 349
+    },
+    {
+      "epoch": 1.429887410440123,
+      "grad_norm": 1.0308408737182617,
+      "learning_rate": 1.1382867151647333e-05,
+      "loss": 0.1485,
+      "step": 350
+    },
+    {
+      "epoch": 1.4339815762538382,
+      "grad_norm": 0.9077311158180237,
+      "learning_rate": 1.1338531386752618e-05,
+      "loss": 0.1472,
+      "step": 351
+    },
+    {
+      "epoch": 1.4380757420675536,
+      "grad_norm": 1.0163536071777344,
+      "learning_rate": 1.1294168814540554e-05,
+      "loss": 0.169,
+      "step": 352
+    },
+    {
+      "epoch": 1.442169907881269,
+      "grad_norm": 1.095957636833191,
+      "learning_rate": 1.1249780323478585e-05,
+      "loss": 0.1687,
+      "step": 353
+    },
+    {
+      "epoch": 1.4462640736949846,
+      "grad_norm": 1.0116571187973022,
+      "learning_rate": 1.1205366802553231e-05,
+      "loss": 0.1687,
+      "step": 354
+    },
+    {
+      "epoch": 1.4503582395087,
+      "grad_norm": 0.893722653388977,
+      "learning_rate": 1.1160929141252303e-05,
+      "loss": 0.1483,
+      "step": 355
+    },
+    {
+      "epoch": 1.4544524053224155,
+      "grad_norm": 1.1401066780090332,
+      "learning_rate": 1.1116468229547079e-05,
+      "loss": 0.171,
+      "step": 356
+    },
+    {
+      "epoch": 1.458546571136131,
+      "grad_norm": 0.9569492340087891,
+      "learning_rate": 1.107198495787448e-05,
+      "loss": 0.1573,
+      "step": 357
+    },
+    {
+      "epoch": 1.4626407369498464,
+      "grad_norm": 0.9641392230987549,
+      "learning_rate": 1.1027480217119245e-05,
+      "loss": 0.1518,
+      "step": 358
+    },
+    {
+      "epoch": 1.4667349027635619,
+      "grad_norm": 0.986418604850769,
+      "learning_rate": 1.0982954898596072e-05,
+      "loss": 0.1637,
+      "step": 359
+    },
+    {
+      "epoch": 1.4708290685772774,
+      "grad_norm": 0.9645577073097229,
+      "learning_rate": 1.0938409894031793e-05,
+      "loss": 0.1398,
+      "step": 360
+    },
+    {
+      "epoch": 1.4749232343909928,
+      "grad_norm": 1.0230332612991333,
+      "learning_rate": 1.0893846095547493e-05,
+      "loss": 0.1491,
+      "step": 361
+    },
+    {
+      "epoch": 1.4790174002047083,
+      "grad_norm": 0.9442994594573975,
+      "learning_rate": 1.084926439564065e-05,
+      "loss": 0.1593,
+      "step": 362
+    },
+    {
+      "epoch": 1.4831115660184238,
+      "grad_norm": 0.8560605645179749,
+      "learning_rate": 1.0804665687167262e-05,
+      "loss": 0.1594,
+      "step": 363
+    },
+    {
+      "epoch": 1.4872057318321392,
+      "grad_norm": 1.018290400505066,
+      "learning_rate": 1.0760050863323961e-05,
+      "loss": 0.1604,
+      "step": 364
+    },
+    {
+      "epoch": 1.4912998976458547,
+      "grad_norm": 0.8661187291145325,
+      "learning_rate": 1.0715420817630137e-05,
+      "loss": 0.1432,
+      "step": 365
+    },
+    {
+      "epoch": 1.4953940634595702,
+      "grad_norm": 0.9243437647819519,
+      "learning_rate": 1.0670776443910024e-05,
+      "loss": 0.151,
+      "step": 366
+    },
+    {
+      "epoch": 1.4994882292732856,
+      "grad_norm": 0.9546313285827637,
+      "learning_rate": 1.062611863627482e-05,
+      "loss": 0.1519,
+      "step": 367
+    },
+    {
+      "epoch": 1.503582395087001,
+      "grad_norm": 0.9470425248146057,
+      "learning_rate": 1.0581448289104759e-05,
+      "loss": 0.1463,
+      "step": 368
+    },
+    {
+      "epoch": 1.5076765609007166,
+      "grad_norm": 1.0364755392074585,
+      "learning_rate": 1.0536766297031216e-05,
+      "loss": 0.1695,
+      "step": 369
+    },
+    {
+      "epoch": 1.511770726714432,
+      "grad_norm": 0.9163882732391357,
+      "learning_rate": 1.0492073554918782e-05,
+      "loss": 0.1549,
+      "step": 370
+    },
+    {
+      "epoch": 1.5158648925281475,
+      "grad_norm": 1.1011639833450317,
+      "learning_rate": 1.0447370957847343e-05,
+      "loss": 0.154,
+      "step": 371
+    },
+    {
+      "epoch": 1.519959058341863,
+      "grad_norm": 0.8807597160339355,
+      "learning_rate": 1.0402659401094154e-05,
+      "loss": 0.1364,
+      "step": 372
+    },
+    {
+      "epoch": 1.5240532241555784,
+      "grad_norm": 0.9387779831886292,
+      "learning_rate": 1.0357939780115906e-05,
+      "loss": 0.1593,
+      "step": 373
+    },
+    {
+      "epoch": 1.528147389969294,
+      "grad_norm": 1.1142334938049316,
+      "learning_rate": 1.0313212990530804e-05,
+      "loss": 0.1547,
+      "step": 374
+    },
+    {
+      "epoch": 1.5322415557830094,
+      "grad_norm": 1.0406149625778198,
+      "learning_rate": 1.0268479928100615e-05,
+      "loss": 0.1639,
+      "step": 375
+    },
+    {
+      "epoch": 1.5363357215967248,
+      "grad_norm": 0.8890196084976196,
+      "learning_rate": 1.0223741488712732e-05,
+      "loss": 0.1357,
+      "step": 376
+    },
+    {
+      "epoch": 1.54042988741044,
+      "grad_norm": 0.8725599050521851,
+      "learning_rate": 1.0178998568362243e-05,
+      "loss": 0.1334,
+      "step": 377
+    },
+    {
+      "epoch": 1.5445240532241555,
+      "grad_norm": 1.0039124488830566,
+      "learning_rate": 1.0134252063133976e-05,
+      "loss": 0.1544,
+      "step": 378
+    },
+    {
+      "epoch": 1.548618219037871,
+      "grad_norm": 0.9957043528556824,
+      "learning_rate": 1.0089502869184549e-05,
+      "loss": 0.1578,
+      "step": 379
+    },
+    {
+      "epoch": 1.5527123848515865,
+      "grad_norm": 0.8412466049194336,
+      "learning_rate": 1.0044751882724436e-05,
+      "loss": 0.1328,
+      "step": 380
+    },
+    {
+      "epoch": 1.556806550665302,
+      "grad_norm": 0.9724802374839783,
+      "learning_rate": 1e-05,
+      "loss": 0.1494,
+      "step": 381
+    },
+    {
+      "epoch": 1.5609007164790174,
+      "grad_norm": 0.8915446996688843,
+      "learning_rate": 9.955248117275566e-06,
+      "loss": 0.1528,
+      "step": 382
+    },
+    {
+      "epoch": 1.5649948822927329,
+      "grad_norm": 0.8682056069374084,
+      "learning_rate": 9.910497130815454e-06,
+      "loss": 0.1378,
+      "step": 383
+    },
+    {
+      "epoch": 1.5690890481064483,
+      "grad_norm": 0.9853907227516174,
+      "learning_rate": 9.865747936866027e-06,
+      "loss": 0.1521,
+      "step": 384
+    },
+    {
+      "epoch": 1.5731832139201638,
+      "grad_norm": 1.1003942489624023,
+      "learning_rate": 9.821001431637759e-06,
+      "loss": 0.1592,
+      "step": 385
+    },
+    {
+      "epoch": 1.5772773797338793,
+      "grad_norm": 0.9457170367240906,
+      "learning_rate": 9.776258511287271e-06,
+      "loss": 0.1404,
+      "step": 386
+    },
+    {
+      "epoch": 1.5813715455475945,
+      "grad_norm": 1.0267016887664795,
+      "learning_rate": 9.73152007189939e-06,
+      "loss": 0.1631,
+      "step": 387
+    },
+    {
+      "epoch": 1.58546571136131,
+      "grad_norm": 0.899929940700531,
+      "learning_rate": 9.6867870094692e-06,
+      "loss": 0.1596,
+      "step": 388
+    },
+    {
+      "epoch": 1.5895598771750254,
+      "grad_norm": 0.904625415802002,
+      "learning_rate": 9.642060219884096e-06,
+      "loss": 0.1431,
+      "step": 389
+    },
+    {
+      "epoch": 1.593654042988741,
+      "grad_norm": 1.0200403928756714,
+      "learning_rate": 9.597340598905851e-06,
+      "loss": 0.1475,
+      "step": 390
+    },
+    {
+      "epoch": 1.5977482088024564,
+      "grad_norm": 0.8861089944839478,
+      "learning_rate": 9.55262904215266e-06,
+      "loss": 0.1495,
+      "step": 391
+    },
+    {
+      "epoch": 1.6018423746161718,
+      "grad_norm": 0.9779448509216309,
+      "learning_rate": 9.50792644508122e-06,
+      "loss": 0.1514,
+      "step": 392
+    },
+    {
+      "epoch": 1.6059365404298873,
+      "grad_norm": 0.8129394054412842,
+      "learning_rate": 9.463233702968784e-06,
+      "loss": 0.1482,
+      "step": 393
+    },
+    {
+      "epoch": 1.6100307062436028,
+      "grad_norm": 0.8793542385101318,
+      "learning_rate": 9.418551710895243e-06,
+      "loss": 0.1505,
+      "step": 394
+    },
+    {
+      "epoch": 1.6141248720573182,
+      "grad_norm": 0.9809761643409729,
+      "learning_rate": 9.373881363725182e-06,
+      "loss": 0.1548,
+      "step": 395
+    },
+    {
+      "epoch": 1.6182190378710337,
+      "grad_norm": 0.9904576539993286,
+      "learning_rate": 9.329223556089976e-06,
+      "loss": 0.1579,
+      "step": 396
+    },
+    {
+      "epoch": 1.6223132036847492,
+      "grad_norm": 0.8856755495071411,
+      "learning_rate": 9.284579182369868e-06,
+      "loss": 0.1524,
+      "step": 397
+    },
+    {
+      "epoch": 1.6264073694984647,
+      "grad_norm": 1.0314946174621582,
+      "learning_rate": 9.239949136676042e-06,
+      "loss": 0.162,
+      "step": 398
+    },
+    {
+      "epoch": 1.6305015353121801,
+      "grad_norm": 0.9784588813781738,
+      "learning_rate": 9.195334312832742e-06,
+      "loss": 0.1463,
+      "step": 399
+    },
+    {
+      "epoch": 1.6345957011258956,
+      "grad_norm": 0.8726085424423218,
+      "learning_rate": 9.15073560435935e-06,
+      "loss": 0.1324,
+      "step": 400
+    },
+    {
+      "epoch": 1.638689866939611,
+      "grad_norm": 0.9043903946876526,
+      "learning_rate": 9.10615390445251e-06,
+      "loss": 0.1482,
+      "step": 401
+    },
+    {
+      "epoch": 1.6427840327533265,
+      "grad_norm": 0.9523571729660034,
+      "learning_rate": 9.061590105968208e-06,
+      "loss": 0.1404,
+      "step": 402
+    },
+    {
+      "epoch": 1.646878198567042,
+      "grad_norm": 0.9627721905708313,
+      "learning_rate": 9.01704510140393e-06,
+      "loss": 0.1367,
+      "step": 403
+    },
+    {
+      "epoch": 1.6509723643807575,
+      "grad_norm": 0.9449582695960999,
+      "learning_rate": 8.97251978288076e-06,
+      "loss": 0.1557,
+      "step": 404
+    },
+    {
+      "epoch": 1.655066530194473,
+      "grad_norm": 1.0399394035339355,
+      "learning_rate": 8.928015042125523e-06,
+      "loss": 0.153,
+      "step": 405
+    },
+    {
+      "epoch": 1.6591606960081884,
+      "grad_norm": 0.9468371868133545,
+      "learning_rate": 8.883531770452924e-06,
+      "loss": 0.1544,
+      "step": 406
+    },
+    {
+      "epoch": 1.6632548618219039,
+      "grad_norm": 0.9178614020347595,
+      "learning_rate": 8.839070858747697e-06,
+      "loss": 0.1451,
+      "step": 407
+    },
+    {
+      "epoch": 1.6673490276356193,
+      "grad_norm": 0.8932563662528992,
+      "learning_rate": 8.79463319744677e-06,
+      "loss": 0.1495,
+      "step": 408
+    },
+    {
+      "epoch": 1.6714431934493348,
+      "grad_norm": 0.9476360082626343,
+      "learning_rate": 8.750219676521417e-06,
+      "loss": 0.1402,
+      "step": 409
+    },
+    {
+      "epoch": 1.6755373592630503,
+      "grad_norm": 1.1400271654129028,
+      "learning_rate": 8.705831185459446e-06,
+      "loss": 0.1605,
+      "step": 410
+    },
+    {
+      "epoch": 1.6755373592630503,
+      "eval_loss": 0.2577343285083771,
+      "eval_runtime": 5.9021,
+      "eval_samples_per_second": 13.385,
+      "eval_steps_per_second": 1.694,
+      "step": 410
+    },
+    {
+      "epoch": 1.6796315250767657,
+      "grad_norm": 1.0324385166168213,
+      "learning_rate": 8.661468613247387e-06,
+      "loss": 0.155,
+      "step": 411
+    },
+    {
+      "epoch": 1.6837256908904812,
+      "grad_norm": 0.8972249031066895,
+      "learning_rate": 8.617132848352672e-06,
+      "loss": 0.1493,
+      "step": 412
+    },
+    {
+      "epoch": 1.6878198567041967,
+      "grad_norm": 0.9417198896408081,
+      "learning_rate": 8.572824778705858e-06,
+      "loss": 0.1446,
+      "step": 413
+    },
+    {
+      "epoch": 1.6919140225179121,
+      "grad_norm": 0.9958034753799438,
+      "learning_rate": 8.528545291682839e-06,
+      "loss": 0.1725,
+      "step": 414
+    },
+    {
+      "epoch": 1.6960081883316276,
+      "grad_norm": 0.7882540822029114,
+      "learning_rate": 8.484295274087077e-06,
+      "loss": 0.1332,
+      "step": 415
+    },
+    {
+      "epoch": 1.7001023541453428,
+      "grad_norm": 0.9401816725730896,
+      "learning_rate": 8.440075612131823e-06,
+      "loss": 0.1482,
+      "step": 416
+    },
+    {
+      "epoch": 1.7041965199590583,
+      "grad_norm": 0.8681234121322632,
+      "learning_rate": 8.395887191422397e-06,
+      "loss": 0.1563,
+      "step": 417
+    },
+    {
+      "epoch": 1.7082906857727738,
+      "grad_norm": 0.953637957572937,
+      "learning_rate": 8.351730896938438e-06,
+      "loss": 0.1548,
+      "step": 418
+    },
+    {
+      "epoch": 1.7123848515864892,
+      "grad_norm": 0.9913907051086426,
+      "learning_rate": 8.307607613016166e-06,
+      "loss": 0.151,
+      "step": 419
+    },
+    {
+      "epoch": 1.7164790174002047,
+      "grad_norm": 0.9096423983573914,
+      "learning_rate": 8.263518223330698e-06,
+      "loss": 0.1425,
+      "step": 420
+    },
+    {
+      "epoch": 1.7205731832139202,
+      "grad_norm": 0.8913793563842773,
+      "learning_rate": 8.219463610878336e-06,
+      "loss": 0.1445,
+      "step": 421
+    },
+    {
+      "epoch": 1.7246673490276356,
+      "grad_norm": 0.8762878775596619,
+      "learning_rate": 8.175444657958875e-06,
+      "loss": 0.1328,
+      "step": 422
+    },
+    {
+      "epoch": 1.728761514841351,
+      "grad_norm": 0.906721830368042,
+      "learning_rate": 8.131462246157953e-06,
+      "loss": 0.1432,
+      "step": 423
+    },
+    {
+      "epoch": 1.7328556806550666,
+      "grad_norm": 0.9558602571487427,
+      "learning_rate": 8.087517256329376e-06,
+      "loss": 0.1556,
+      "step": 424
+    },
+    {
+      "epoch": 1.736949846468782,
+      "grad_norm": 1.0226850509643555,
+      "learning_rate": 8.043610568577497e-06,
+      "loss": 0.1526,
+      "step": 425
+    },
+    {
+      "epoch": 1.7410440122824973,
+      "grad_norm": 0.8428735136985779,
+      "learning_rate": 7.999743062239557e-06,
+      "loss": 0.1394,
+      "step": 426
+    },
+    {
+      "epoch": 1.7451381780962127,
+      "grad_norm": 0.8992868065834045,
+      "learning_rate": 7.95591561586811e-06,
+      "loss": 0.1467,
+      "step": 427
+    },
+    {
+      "epoch": 1.7492323439099282,
+      "grad_norm": 0.8938621878623962,
+      "learning_rate": 7.912129107213417e-06,
+      "loss": 0.1436,
+      "step": 428
+    },
+    {
+      "epoch": 1.7533265097236437,
+      "grad_norm": 0.9230063557624817,
+      "learning_rate": 7.868384413205842e-06,
+      "loss": 0.1428,
+      "step": 429
+    },
+    {
+      "epoch": 1.7574206755373591,
+      "grad_norm": 1.0105928182601929,
+      "learning_rate": 7.824682409938328e-06,
+      "loss": 0.1624,
+      "step": 430
+    },
+    {
+      "epoch": 1.7615148413510746,
+      "grad_norm": 0.9786428213119507,
+      "learning_rate": 7.781023972648826e-06,
+      "loss": 0.1508,
+      "step": 431
+    },
+    {
+      "epoch": 1.76560900716479,
+      "grad_norm": 0.9351881742477417,
+      "learning_rate": 7.73740997570278e-06,
+      "loss": 0.1473,
+      "step": 432
+    },
+    {
+      "epoch": 1.7697031729785055,
+      "grad_norm": 0.9618210196495056,
+      "learning_rate": 7.6938412925756e-06,
+      "loss": 0.1374,
+      "step": 433
+    },
+    {
+      "epoch": 1.773797338792221,
+      "grad_norm": 0.9136247634887695,
+      "learning_rate": 7.650318795835179e-06,
+      "loss": 0.1457,
+      "step": 434
+    },
+    {
+      "epoch": 1.7778915046059365,
+      "grad_norm": 1.2237149477005005,
+      "learning_rate": 7.606843357124426e-06,
+      "loss": 0.1448,
+      "step": 435
+    },
+    {
+      "epoch": 1.781985670419652,
+      "grad_norm": 1.0644303560256958,
+      "learning_rate": 7.563415847143782e-06,
+      "loss": 0.1618,
+      "step": 436
+    },
+    {
+      "epoch": 1.7860798362333674,
+      "grad_norm": 0.992850661277771,
+      "learning_rate": 7.520037135633817e-06,
+      "loss": 0.1432,
+      "step": 437
+    },
+    {
+      "epoch": 1.7901740020470829,
+      "grad_norm": 0.8789179921150208,
+      "learning_rate": 7.476708091357783e-06,
+      "loss": 0.1593,
+      "step": 438
+    },
+    {
+      "epoch": 1.7942681678607983,
+      "grad_norm": 0.8858450055122375,
+      "learning_rate": 7.433429582084233e-06,
+      "loss": 0.1421,
+      "step": 439
+    },
+    {
+      "epoch": 1.7983623336745138,
+      "grad_norm": 0.8538363575935364,
+      "learning_rate": 7.39020247456963e-06,
+      "loss": 0.1368,
+      "step": 440
+    },
+    {
+      "epoch": 1.8024564994882293,
+      "grad_norm": 0.9794362187385559,
+      "learning_rate": 7.347027634540993e-06,
+      "loss": 0.1568,
+      "step": 441
+    },
+    {
+      "epoch": 1.8065506653019447,
+      "grad_norm": 0.9608176350593567,
+      "learning_rate": 7.303905926678565e-06,
+      "loss": 0.1736,
+      "step": 442
+    },
+    {
+      "epoch": 1.8106448311156602,
+      "grad_norm": 0.9341726303100586,
+      "learning_rate": 7.260838214598475e-06,
+      "loss": 0.1559,
+      "step": 443
+    },
+    {
+      "epoch": 1.8147389969293757,
+      "grad_norm": 0.9113818407058716,
+      "learning_rate": 7.217825360835475e-06,
+      "loss": 0.154,
+      "step": 444
+    },
+    {
+      "epoch": 1.8188331627430911,
+      "grad_norm": 0.8818947076797485,
+      "learning_rate": 7.174868226825631e-06,
+      "loss": 0.1391,
+      "step": 445
+    },
+    {
+      "epoch": 1.8229273285568066,
+      "grad_norm": 0.8737187385559082,
+      "learning_rate": 7.131967672889101e-06,
+      "loss": 0.1508,
+      "step": 446
+    },
+    {
+      "epoch": 1.827021494370522,
+      "grad_norm": 0.9715389609336853,
+      "learning_rate": 7.089124558212872e-06,
+      "loss": 0.1368,
+      "step": 447
+    },
+    {
+      "epoch": 1.8311156601842375,
+      "grad_norm": 0.936271607875824,
+      "learning_rate": 7.04633974083359e-06,
+      "loss": 0.1468,
+      "step": 448
+    },
+    {
+      "epoch": 1.835209825997953,
+      "grad_norm": 0.8608719110488892,
+      "learning_rate": 7.003614077620348e-06,
+      "loss": 0.1405,
+      "step": 449
+    },
+    {
+      "epoch": 1.8393039918116685,
+      "grad_norm": 1.0865558385849,
+      "learning_rate": 6.960948424257532e-06,
+      "loss": 0.141,
+      "step": 450
+    },
+    {
+      "epoch": 1.843398157625384,
+      "grad_norm": 0.8026862740516663,
+      "learning_rate": 6.918343635227694e-06,
+      "loss": 0.1264,
+      "step": 451
+    },
+    {
+      "epoch": 1.8474923234390994,
+      "grad_norm": 0.9582161903381348,
+      "learning_rate": 6.8758005637944245e-06,
+      "loss": 0.1599,
+      "step": 452
+    },
+    {
+      "epoch": 1.8515864892528149,
+      "grad_norm": 1.0018301010131836,
+      "learning_rate": 6.833320061985278e-06,
+      "loss": 0.155,
+      "step": 453
+    },
+    {
+      "epoch": 1.8556806550665303,
+      "grad_norm": 0.9807232618331909,
+      "learning_rate": 6.7909029805746855e-06,
+      "loss": 0.1519,
+      "step": 454
+    },
+    {
+      "epoch": 1.8597748208802458,
+      "grad_norm": 0.8871752023696899,
+      "learning_rate": 6.7485501690669495e-06,
+      "loss": 0.1231,
+      "step": 455
+    },
+    {
+      "epoch": 1.863868986693961,
+      "grad_norm": 0.9024534225463867,
+      "learning_rate": 6.706262475679205e-06,
+      "loss": 0.1472,
+      "step": 456
+    },
+    {
+      "epoch": 1.8679631525076765,
+      "grad_norm": 1.037083387374878,
+      "learning_rate": 6.664040747324437e-06,
+      "loss": 0.149,
+      "step": 457
+    },
+    {
+      "epoch": 1.872057318321392,
+      "grad_norm": 1.0350216627120972,
+      "learning_rate": 6.62188582959453e-06,
+      "loss": 0.151,
+      "step": 458
+    },
+    {
+      "epoch": 1.8761514841351075,
+      "grad_norm": 0.8358607292175293,
+      "learning_rate": 6.579798566743314e-06,
+      "loss": 0.1331,
+      "step": 459
+    },
+    {
+      "epoch": 1.880245649948823,
+      "grad_norm": 0.9464243054389954,
+      "learning_rate": 6.537779801669677e-06,
+      "loss": 0.1338,
+      "step": 460
+    },
+    {
+      "epoch": 1.8843398157625384,
+      "grad_norm": 0.9431869983673096,
+      "learning_rate": 6.495830375900665e-06,
+      "loss": 0.1404,
+      "step": 461
+    },
+    {
+      "epoch": 1.8884339815762539,
+      "grad_norm": 0.9446228742599487,
+      "learning_rate": 6.453951129574644e-06,
+      "loss": 0.1598,
+      "step": 462
+    },
+    {
+      "epoch": 1.8925281473899693,
+      "grad_norm": 0.9133023023605347,
+      "learning_rate": 6.41214290142447e-06,
+      "loss": 0.1415,
+      "step": 463
+    },
+    {
+      "epoch": 1.8966223132036848,
+      "grad_norm": 0.8346714377403259,
+      "learning_rate": 6.370406528760675e-06,
+      "loss": 0.1384,
+      "step": 464
+    },
+    {
+      "epoch": 1.9007164790174,
+      "grad_norm": 0.9194797277450562,
+      "learning_rate": 6.3287428474547256e-06,
+      "loss": 0.1451,
+      "step": 465
+    },
+    {
+      "epoch": 1.9048106448311155,
+      "grad_norm": 1.0291615724563599,
+      "learning_rate": 6.287152691922264e-06,
+      "loss": 0.143,
+      "step": 466
+    },
+    {
+      "epoch": 1.908904810644831,
+      "grad_norm": 0.9120848178863525,
+      "learning_rate": 6.245636895106403e-06,
+      "loss": 0.1471,
+      "step": 467
+    },
+    {
+      "epoch": 1.9129989764585464,
+      "grad_norm": 0.9780627489089966,
+      "learning_rate": 6.204196288461037e-06,
+      "loss": 0.1532,
+      "step": 468
+    },
+    {
+      "epoch": 1.917093142272262,
+      "grad_norm": 0.9170928597450256,
+      "learning_rate": 6.162831701934203e-06,
+      "loss": 0.1428,
+      "step": 469
+    },
+    {
+      "epoch": 1.9211873080859774,
+      "grad_norm": 0.8668283224105835,
+      "learning_rate": 6.121543963951453e-06,
+      "loss": 0.1375,
+      "step": 470
+    },
+    {
+      "epoch": 1.9252814738996928,
+      "grad_norm": 0.9142316579818726,
+      "learning_rate": 6.080333901399252e-06,
+      "loss": 0.1376,
+      "step": 471
+    },
+    {
+      "epoch": 1.9293756397134083,
+      "grad_norm": 0.9596243500709534,
+      "learning_rate": 6.039202339608432e-06,
+      "loss": 0.1519,
+      "step": 472
+    },
+    {
+      "epoch": 1.9334698055271238,
+      "grad_norm": 0.8200681209564209,
+      "learning_rate": 5.998150102337665e-06,
+      "loss": 0.1294,
+      "step": 473
+    },
+    {
+      "epoch": 1.9375639713408392,
+      "grad_norm": 0.9270857572555542,
+      "learning_rate": 5.957178011756952e-06,
+      "loss": 0.1455,
+      "step": 474
+    },
+    {
+      "epoch": 1.9416581371545547,
+      "grad_norm": 1.033295750617981,
+      "learning_rate": 5.9162868884311596e-06,
+      "loss": 0.1554,
+      "step": 475
+    },
+    {
+      "epoch": 1.9457523029682702,
+      "grad_norm": 0.8937323689460754,
+      "learning_rate": 5.875477551303596e-06,
+      "loss": 0.151,
+      "step": 476
+    },
+    {
+      "epoch": 1.9498464687819856,
+      "grad_norm": 0.9521363377571106,
+      "learning_rate": 5.834750817679606e-06,
+      "loss": 0.1327,
+      "step": 477
+    },
+    {
+      "epoch": 1.953940634595701,
+      "grad_norm": 0.9314703941345215,
+      "learning_rate": 5.794107503210187e-06,
+      "loss": 0.1425,
+      "step": 478
+    },
+    {
+      "epoch": 1.9580348004094166,
+      "grad_norm": 0.975556492805481,
+      "learning_rate": 5.753548421875686e-06,
+      "loss": 0.1439,
+      "step": 479
+    },
+    {
+      "epoch": 1.962128966223132,
+      "grad_norm": 0.952316164970398,
+      "learning_rate": 5.713074385969457e-06,
+      "loss": 0.1403,
+      "step": 480
+    },
+    {
+      "epoch": 1.9662231320368475,
+      "grad_norm": 0.8896644115447998,
+      "learning_rate": 5.672686206081638e-06,
+      "loss": 0.1353,
+      "step": 481
+    },
+    {
+      "epoch": 1.970317297850563,
+      "grad_norm": 0.9095272421836853,
+      "learning_rate": 5.632384691082874e-06,
+      "loss": 0.137,
+      "step": 482
+    },
+    {
+      "epoch": 1.9744114636642784,
+      "grad_norm": 0.823320746421814,
+      "learning_rate": 5.5921706481081405e-06,
+      "loss": 0.1366,
+      "step": 483
+    },
+    {
+      "epoch": 1.978505629477994,
+      "grad_norm": 0.8442295789718628,
+      "learning_rate": 5.55204488254059e-06,
+      "loss": 0.1337,
+      "step": 484
+    },
+    {
+      "epoch": 1.9825997952917094,
+      "grad_norm": 0.9687731266021729,
+      "learning_rate": 5.512008197995379e-06,
+      "loss": 0.1487,
+      "step": 485
+    },
+    {
+      "epoch": 1.9866939611054248,
+      "grad_norm": 0.8544232249259949,
+      "learning_rate": 5.47206139630363e-06,
+      "loss": 0.1433,
+      "step": 486
+    },
+    {
+      "epoch": 1.9907881269191403,
+      "grad_norm": 1.0014803409576416,
+      "learning_rate": 5.432205277496327e-06,
+      "loss": 0.1402,
+      "step": 487
+    },
+    {
+      "epoch": 1.9948822927328558,
+      "grad_norm": 0.9409521222114563,
+      "learning_rate": 5.3924406397883174e-06,
+      "loss": 0.1374,
+      "step": 488
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 732,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 244,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 7.874994391608197e+17,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}

3b-w-cot+/checkpoint-488/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:632e641a15180cc852702434a272df94b8012efb84c5e296eb59b1554cdab170
+size 10744

3b-w-cot+/checkpoint-488/zero_to_fp32.py

@@ -0,0 +1,760 @@
+#!/usr/bin/env python
+
+# Copyright (c) Microsoft Corporation.
+# SPDX-License-Identifier: Apache-2.0
+
+# DeepSpeed Team
+
+# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets
+# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in
+# the future. Once extracted, the weights don't require DeepSpeed and can be used in any
+# application.
+#
+# example:
+#   python zero_to_fp32.py . output_dir/
+#   or
+#   python zero_to_fp32.py . output_dir/ --safe_serialization
+
+import argparse
+import torch
+import glob
+import math
+import os
+import re
+import gc
+import json
+import numpy as np
+from tqdm import tqdm
+from collections import OrderedDict
+from dataclasses import dataclass
+
+# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with
+# DeepSpeed data structures it has to be available in the current python environment.
+from deepspeed.utils import logger
+from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS,
+                                            FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES,
+                                            FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS)
+
+
+@dataclass
+class zero_model_state:
+    buffers: dict()
+    param_shapes: dict()
+    shared_params: list
+    ds_version: int
+    frozen_param_shapes: dict()
+    frozen_param_fragments: dict()
+
+
+debug = 0
+
+# load to cpu
+device = torch.device('cpu')
+
+
+def atoi(text):
+    return int(text) if text.isdigit() else text
+
+
+def natural_keys(text):
+    '''
+    alist.sort(key=natural_keys) sorts in human order
+    http://nedbatchelder.com/blog/200712/human_sorting.html
+    (See Toothy's implementation in the comments)
+    '''
+    return [atoi(c) for c in re.split(r'(\d+)', text)]
+
+
+def get_model_state_file(checkpoint_dir, zero_stage):
+    if not os.path.isdir(checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist")
+
+    # there should be only one file
+    if zero_stage <= 2:
+        file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt")
+    elif zero_stage == 3:
+        file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt")
+
+    if not os.path.exists(file):
+        raise FileNotFoundError(f"can't find model states file at '{file}'")
+
+    return file
+
+
+def get_checkpoint_files(checkpoint_dir, glob_pattern):
+    # XXX: need to test that this simple glob rule works for multi-node setup too
+    ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys)
+
+    if len(ckpt_files) == 0:
+        raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'")
+
+    return ckpt_files
+
+
+def get_optim_files(checkpoint_dir):
+    return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt")
+
+
+def get_model_state_files(checkpoint_dir):
+    return get_checkpoint_files(checkpoint_dir, "*_model_states.pt")
+
+
+def parse_model_states(files):
+    zero_model_states = []
+    for file in files:
+        state_dict = torch.load(file, map_location=device, weights_only=False)
+
+        if BUFFER_NAMES not in state_dict:
+            raise ValueError(f"{file} is not a model state checkpoint")
+        buffer_names = state_dict[BUFFER_NAMES]
+        if debug:
+            print("Found buffers:", buffer_names)
+
+        # recover just the buffers while restoring them to fp32 if they were saved in fp16
+        buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names}
+        param_shapes = state_dict[PARAM_SHAPES]
+
+        # collect parameters that are included in param_shapes
+        param_names = []
+        for s in param_shapes:
+            for name in s.keys():
+                param_names.append(name)
+
+        # update with frozen parameters
+        frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None)
+        if frozen_param_shapes is not None:
+            if debug:
+                print(f"Found frozen_param_shapes: {frozen_param_shapes}")
+            param_names += list(frozen_param_shapes.keys())
+
+        # handle shared params
+        shared_params = [[k, v] for k, v in state_dict["shared_params"].items()]
+
+        ds_version = state_dict.get(DS_VERSION, None)
+
+        frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None)
+
+        z_model_state = zero_model_state(buffers=buffers,
+                                         param_shapes=param_shapes,
+                                         shared_params=shared_params,
+                                         ds_version=ds_version,
+                                         frozen_param_shapes=frozen_param_shapes,
+                                         frozen_param_fragments=frozen_param_fragments)
+        zero_model_states.append(z_model_state)
+
+    return zero_model_states
+
+
+def parse_optim_states(files, ds_checkpoint_dir):
+    total_files = len(files)
+    state_dicts = []
+    for f in tqdm(files, desc='Loading checkpoint shards'):
+        state_dict = torch.load(f, map_location=device, mmap=True, weights_only=False)
+        # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights
+        # and also handle the case where it was already removed by another helper script
+        state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None)
+        state_dicts.append(state_dict)
+
+    if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]:
+        raise ValueError(f"{files[0]} is not a zero checkpoint")
+    zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE]
+    world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT]
+
+    # For ZeRO-2 each param group can have different partition_count as data parallelism for expert
+    # parameters can be different from data parallelism for non-expert parameters. So we can just
+    # use the max of the partition_count to get the dp world_size.
+
+    if type(world_size) is list:
+        world_size = max(world_size)
+
+    if world_size != total_files:
+        raise ValueError(
+            f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. "
+            "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes."
+        )
+
+    # the groups are named differently in each stage
+    if zero_stage <= 2:
+        fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS
+    elif zero_stage == 3:
+        fp32_groups_key = FP32_FLAT_GROUPS
+    else:
+        raise ValueError(f"unknown zero stage {zero_stage}")
+
+    fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))]
+    return zero_stage, world_size, fp32_flat_groups
+
+
+def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters):
+    """
+    Returns fp32 state_dict reconstructed from ds checkpoint
+
+    Args:
+        - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are)
+
+    """
+    print(f"Processing zero checkpoint '{ds_checkpoint_dir}'")
+
+    optim_files = get_optim_files(ds_checkpoint_dir)
+    zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir)
+    print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}")
+
+    model_files = get_model_state_files(ds_checkpoint_dir)
+
+    zero_model_states = parse_model_states(model_files)
+    print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}')
+
+    if zero_stage <= 2:
+        return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                                          exclude_frozen_parameters)
+    elif zero_stage == 3:
+        return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                                          exclude_frozen_parameters)
+
+
+def _zero2_merge_frozen_params(state_dict, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+    frozen_param_fragments = zero_model_states[0].frozen_param_fragments
+
+    if debug:
+        num_elem = sum(s.numel() for s in frozen_param_shapes.values())
+        print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in frozen_param_fragments.values()])
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        state_dict[name] = frozen_param_fragments[name]
+
+        if debug:
+            print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _has_callable(obj, fn):
+    attr = getattr(obj, fn, None)
+    return callable(attr)
+
+
+def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+
+    # Reconstruction protocol:
+    #
+    # XXX: document this
+
+    if debug:
+        for i in range(world_size):
+            for j in range(len(fp32_flat_groups[0])):
+                print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}")
+
+    # XXX: memory usage doubles here (zero2)
+    num_param_groups = len(fp32_flat_groups[0])
+    merged_single_partition_of_fp32_groups = []
+    for i in range(num_param_groups):
+        merged_partitions = [sd[i] for sd in fp32_flat_groups]
+        full_single_fp32_vector = torch.cat(merged_partitions, 0)
+        merged_single_partition_of_fp32_groups.append(full_single_fp32_vector)
+    avail_numel = sum(
+        [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups])
+
+    if debug:
+        wanted_params = sum([len(shapes) for shapes in param_shapes])
+        wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes])
+        # not asserting if there is a mismatch due to possible padding
+        print(f"Have {avail_numel} numels to process.")
+        print(f"Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    total_numel = 0
+    total_params = 0
+    for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups):
+        offset = 0
+        avail_numel = full_single_fp32_vector.numel()
+        for name, shape in shapes.items():
+
+            unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape)
+            total_numel += unpartitioned_numel
+            total_params += 1
+
+            if debug:
+                print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+            state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape)
+            offset += unpartitioned_numel
+
+        # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and
+        # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex
+        # paddings performed in the code it's almost impossible to predict the exact numbers w/o the
+        # live optimizer object, so we are checking that the numbers are within the right range
+        align_to = 2 * world_size
+
+        def zero2_align(x):
+            return align_to * math.ceil(x / align_to)
+
+        if debug:
+            print(f"original offset={offset}, avail_numel={avail_numel}")
+
+        offset = zero2_align(offset)
+        avail_numel = zero2_align(avail_numel)
+
+        if debug:
+            print(f"aligned  offset={offset}, avail_numel={avail_numel}")
+
+        # Sanity check
+        if offset != avail_numel:
+            raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+    print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                               exclude_frozen_parameters):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    if not exclude_frozen_parameters:
+        _zero2_merge_frozen_params(state_dict, zero_model_states)
+
+    _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def zero3_partitioned_param_info(unpartitioned_numel, world_size):
+    remainder = unpartitioned_numel % world_size
+    padding_numel = (world_size - remainder) if remainder else 0
+    partitioned_numel = math.ceil(unpartitioned_numel / world_size)
+    return partitioned_numel, padding_numel
+
+
+def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    if debug:
+        for i in range(world_size):
+            num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values())
+            print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in zero_model_states[0].frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states)
+        state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape)
+
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+class GatheredTensor:
+    """
+    A pseudo tensor that collects partitioned weights.
+    It is more memory efficient when there are multiple groups.
+    """
+
+    def __init__(self, flat_groups, flat_groups_offset, offset, partitioned_numel, shape):
+        self.flat_groups = flat_groups
+        self.flat_groups_offset = flat_groups_offset
+        self.offset = offset
+        self.partitioned_numel = partitioned_numel
+        self.shape = shape
+        self.dtype = self.flat_groups[0][0].dtype
+
+    def contiguous(self):
+        """
+        Merge partitioned weights from flat_groups into a single tensor.
+        """
+        end_idx = self.offset + self.partitioned_numel
+        world_size = len(self.flat_groups)
+        pad_flat_param_chunks = []
+
+        for rank_i in range(world_size):
+            # for each rank, we need to collect weights from related group/groups
+            flat_groups_at_rank_i = self.flat_groups[rank_i]
+            start_group_id = None
+            end_group_id = None
+            for group_id in range(len(self.flat_groups_offset)):
+                if self.flat_groups_offset[group_id] <= self.offset < self.flat_groups_offset[group_id + 1]:
+                    start_group_id = group_id
+                if self.flat_groups_offset[group_id] < end_idx <= self.flat_groups_offset[group_id + 1]:
+                    end_group_id = group_id
+                    break
+            # collect weights from related group/groups
+            for group_id in range(start_group_id, end_group_id + 1):
+                flat_tensor = flat_groups_at_rank_i[group_id]
+                start_offset = self.offset - self.flat_groups_offset[group_id]
+                end_offset = min(end_idx, self.flat_groups_offset[group_id + 1]) - self.flat_groups_offset[group_id]
+                pad_flat_param_chunks.append(flat_tensor[start_offset:end_offset])
+
+        # collect weights from all ranks
+        pad_flat_param = torch.cat(pad_flat_param_chunks, dim=0)
+        param = pad_flat_param[:self.shape.numel()].view(self.shape).contiguous()
+        return param
+
+
+def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+    avail_numel = sum([flat_group.numel() for flat_group in fp32_flat_groups[0]]) * world_size
+
+    # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each
+    # param, re-consolidating each param, while dealing with padding if any
+
+    # merge list of dicts, preserving order
+    param_shapes = {k: v for d in param_shapes for k, v in d.items()}
+
+    if debug:
+        for i in range(world_size):
+            print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}")
+
+        wanted_params = len(param_shapes)
+        wanted_numel = sum(shape.numel() for shape in param_shapes.values())
+        # not asserting if there is a mismatch due to possible padding
+        avail_numel = fp32_flat_groups[0].numel() * world_size
+        print(f"Trainable params: Have {avail_numel} numels to process.")
+        print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    offset = 0
+    total_numel = 0
+    total_params = 0
+    flat_groups_offset = [0] + list(np.cumsum([flat_tensor.numel() for flat_tensor in fp32_flat_groups[0]]))
+    for name, shape in tqdm(param_shapes.items(), desc='Gathering sharded weights'):
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+        total_params += 1
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+        # memory efficient tensor
+        tensor = GatheredTensor(fp32_flat_groups, flat_groups_offset, offset, partitioned_numel, shape)
+        state_dict[name] = tensor
+        offset += partitioned_numel
+
+    offset *= world_size
+
+    # Sanity check
+    if offset != avail_numel:
+        raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+    print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                               exclude_frozen_parameters):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    if not exclude_frozen_parameters:
+        _zero3_merge_frozen_params(state_dict, world_size, zero_model_states)
+
+    _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def to_torch_tensor(state_dict, return_empty_tensor=False):
+    """
+    Convert state_dict of GatheredTensor to torch tensor
+    """
+    torch_state_dict = {}
+    converted_tensors = {}
+    for name, tensor in state_dict.items():
+        tensor_id = id(tensor)
+        if tensor_id in converted_tensors:  # shared tensors
+            shared_tensor = torch_state_dict[converted_tensors[tensor_id]]
+            torch_state_dict[name] = shared_tensor
+        else:
+            converted_tensors[tensor_id] = name
+            if return_empty_tensor:
+                torch_state_dict[name] = torch.empty(tensor.shape, dtype=tensor.dtype)
+            else:
+                torch_state_dict[name] = tensor.contiguous()
+    return torch_state_dict
+
+
+def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir,
+                                             tag=None,
+                                             exclude_frozen_parameters=False,
+                                             lazy_mode=False):
+    """
+    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with
+    ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example
+    via a model hub.
+
+    Args:
+        - ``checkpoint_dir``: path to the desired checkpoint folder
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14``
+        - ``exclude_frozen_parameters``: exclude frozen parameters
+        - ``lazy_mode``: get state_dict in lazy mode. It returns a dict of pesduo tensor instead of torch tensor, which is more memory efficient.
+          Convert the pesduo tensor to torch tensor by ``.contiguous()``
+
+    Returns:
+        - pytorch ``state_dict``
+
+    A typical usage might be ::
+
+        from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
+        # do the training and checkpoint saving
+        state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu
+        model = model.cpu() # move to cpu
+        model.load_state_dict(state_dict)
+        # submit to model hub or save the model to share with others
+
+    In this example the ``model`` will no longer be usable in the deepspeed context of the same
+    application. i.e. you will need to re-initialize the deepspeed engine, since
+    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+    If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead.
+
+    Note: the above usage may not work if your application doesn't have sufficient free CPU memory.
+    You may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with
+    the checkpoint. Or you can load state_dict in lazy mode ::
+
+        from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
+        state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, lazy_mode=True) # not on cpu
+        for name, lazy_tensor in state_dict.item():
+            tensor = lazy_tensor.contiguous()  # to cpu
+            print(name, tensor)
+            # del tensor to release memory if it no longer in use
+    """
+    if tag is None:
+        latest_path = os.path.join(checkpoint_dir, 'latest')
+        if os.path.isfile(latest_path):
+            with open(latest_path, 'r') as fd:
+                tag = fd.read().strip()
+        else:
+            raise ValueError(f"Unable to find 'latest' file at {latest_path}")
+
+    ds_checkpoint_dir = os.path.join(checkpoint_dir, tag)
+
+    if not os.path.isdir(ds_checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist")
+
+    state_dict = _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters)
+    if lazy_mode:
+        return state_dict
+    else:
+        return to_torch_tensor(state_dict)
+
+
+def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir,
+                                               output_dir,
+                                               max_shard_size="5GB",
+                                               safe_serialization=False,
+                                               tag=None,
+                                               exclude_frozen_parameters=False):
+    """
+    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be
+    loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed.
+
+    Args:
+        - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+        - ``output_dir``: directory to the pytorch fp32 state_dict output files
+        - ``max_shard_size``: the maximum size for a checkpoint before being sharded, default value is 5GB
+        - ``safe_serialization``:  whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+        - ``exclude_frozen_parameters``: exclude frozen parameters
+    """
+
+    # Dependency pre-check
+    if safe_serialization:
+        try:
+            from safetensors.torch import save_file
+        except ImportError:
+            print('If you want to use `safe_serialization`, please `pip install safetensors`')
+            raise
+    if max_shard_size is not None:
+        try:
+            from huggingface_hub import split_torch_state_dict_into_shards
+        except ImportError:
+            print('If you want to use `max_shard_size`, please `pip install huggingface_hub`')
+            raise
+
+    # Convert zero checkpoint to state_dict
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir,
+                                                          tag,
+                                                          exclude_frozen_parameters,
+                                                          lazy_mode=True)
+
+    # Shard the model if it is too big.
+    weights_name = "model.safetensors" if safe_serialization else "pytorch_model.bin"
+    if max_shard_size is not None:
+        filename_pattern = weights_name.replace(".bin", "{suffix}.bin").replace(".safetensors", "{suffix}.safetensors")
+        # an memory-efficient approach for sharding
+        empty_state_dict = to_torch_tensor(state_dict, return_empty_tensor=True)
+        state_dict_split = split_torch_state_dict_into_shards(empty_state_dict,
+                                                              filename_pattern=filename_pattern,
+                                                              max_shard_size=max_shard_size)
+    else:
+        from collections import namedtuple
+        StateDictSplit = namedtuple("StateDictSplit", ["is_sharded", "filename_to_tensors"])
+        state_dict_split = StateDictSplit(is_sharded=False,
+                                          filename_to_tensors={weights_name: list(state_dict.keys())})
+
+    # Save the model by shard
+    os.makedirs(output_dir, exist_ok=True)
+    filename_to_tensors = state_dict_split.filename_to_tensors.items()
+    for shard_file, tensors in tqdm(filename_to_tensors, desc="Saving checkpoint shards"):
+        shard_state_dict = {tensor_name: state_dict[tensor_name] for tensor_name in tensors}
+        shard_state_dict = to_torch_tensor(shard_state_dict)
+        output_path = os.path.join(output_dir, shard_file)
+        if safe_serialization:
+            save_file(shard_state_dict, output_path, metadata={"format": "pt"})
+        else:
+            torch.save(shard_state_dict, output_path)
+        # release the memory of current shard
+        for tensor_name in list(shard_state_dict.keys()):
+            del state_dict[tensor_name]
+            del shard_state_dict[tensor_name]
+        del shard_state_dict
+        gc.collect()
+
+    # Save index if sharded
+    if state_dict_split.is_sharded:
+        index = {
+            "metadata": state_dict_split.metadata,
+            "weight_map": state_dict_split.tensor_to_filename,
+        }
+        save_index_file = "model.safetensors.index.json" if safe_serialization else "pytorch_model.bin.index.json"
+        save_index_file = os.path.join(output_dir, save_index_file)
+        with open(save_index_file, "w", encoding="utf-8") as f:
+            content = json.dumps(index, indent=2, sort_keys=True) + "\n"
+            f.write(content)
+
+
+def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None):
+    """
+    1. Put the provided model to cpu
+    2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict``
+    3. Load it into the provided model
+
+    Args:
+        - ``model``: the model object to update
+        - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+
+    Returns:
+        - ``model`: modified model
+
+    Make sure you have plenty of CPU memory available before you call this function. If you don't
+    have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it
+    conveniently placed for you in the checkpoint folder.
+
+    A typical usage might be ::
+
+        from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint
+        model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir)
+        # submit to model hub or save the model to share with others
+
+    Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context
+    of the same application. i.e. you will need to re-initialize the deepspeed engine, since
+    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+    """
+    logger.info(f"Extracting fp32 weights")
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag)
+
+    logger.info(f"Overwriting model with fp32 weights")
+    model = model.cpu()
+    model.load_state_dict(state_dict, strict=False)
+
+    return model
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser()
+    parser.add_argument("checkpoint_dir",
+                        type=str,
+                        help="path to the desired checkpoint folder, e.g., path/checkpoint-12")
+    parser.add_argument("output_dir",
+                        type=str,
+                        help="directory to the pytorch fp32 state_dict output files"
+                        "(e.g. path/checkpoint-12-output/)")
+    parser.add_argument(
+        "--max_shard_size",
+        type=str,
+        default="5GB",
+        help="The maximum size for a checkpoint before being sharded. Checkpoints shard will then be each of size"
+        "lower than this size. If expressed as a string, needs to be digits followed by a unit (like `5MB`"
+        "We default it to 5GB in order for models to be able to run easily on free-tier google colab instances"
+        "without CPU OOM issues.")
+    parser.add_argument(
+        "--safe_serialization",
+        default=False,
+        action='store_true',
+        help="Whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).")
+    parser.add_argument("-t",
+                        "--tag",
+                        type=str,
+                        default=None,
+                        help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1")
+    parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters")
+    parser.add_argument("-d", "--debug", action='store_true', help="enable debug")
+    args = parser.parse_args()
+
+    debug = args.debug
+
+    convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir,
+                                               args.output_dir,
+                                               max_shard_size=args.max_shard_size,
+                                               safe_serialization=args.safe_serialization,
+                                               tag=args.tag,
+                                               exclude_frozen_parameters=args.exclude_frozen_parameters)

3b-w-cot+/checkpoint-732/added_tokens.json

@@ -0,0 +1,24 @@
+{
+  "</tool_call>": 151658,
+  "<tool_call>": 151657,
+  "<|box_end|>": 151649,
+  "<|box_start|>": 151648,
+  "<|endoftext|>": 151643,
+  "<|file_sep|>": 151664,
+  "<|fim_middle|>": 151660,
+  "<|fim_pad|>": 151662,
+  "<|fim_prefix|>": 151659,
+  "<|fim_suffix|>": 151661,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644,
+  "<|image_pad|>": 151655,
+  "<|object_ref_end|>": 151647,
+  "<|object_ref_start|>": 151646,
+  "<|quad_end|>": 151651,
+  "<|quad_start|>": 151650,
+  "<|repo_name|>": 151663,
+  "<|video_pad|>": 151656,
+  "<|vision_end|>": 151653,
+  "<|vision_pad|>": 151654,
+  "<|vision_start|>": 151652
+}

3b-w-cot+/checkpoint-732/config.json

@@ -0,0 +1,28 @@
+{
+  "_name_or_path": "ckpt/3b-w-cot/checkpoint-747",
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "eos_token_id": 151645,
+  "hidden_act": "silu",
+  "hidden_size": 2048,
+  "initializer_range": 0.02,
+  "intermediate_size": 11008,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 70,
+  "model_type": "qwen2",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 36,
+  "num_key_value_heads": 2,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.49.0",
+  "use_cache": false,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}

3b-w-cot+/checkpoint-732/generation_config.json

@@ -0,0 +1,14 @@
+{
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "repetition_penalty": 1.05,
+  "temperature": 0.7,
+  "top_k": 20,
+  "top_p": 0.8,
+  "transformers_version": "4.49.0"
+}

3b-w-cot+/checkpoint-732/global_step730/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d2dd01b4273d756f834d5ebd3324e6d0d71312aa8d814f35c6dda8790c5006ba
+size 9306058322

3b-w-cot+/checkpoint-732/global_step730/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3e4804f00f9eba5efd889917e4724e26646bac2c402dbc35ec9f1288ebb4160a
+size 9306060690

3b-w-cot+/checkpoint-732/global_step730/mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:530e07a5d4c948fb79b8e599c4a31ff8c53744a24058e20367218f137b22a48e
+size 6171993592